From bruns at 2mbit.com Tue Jan 4 14:44:06 2005 From: bruns at 2mbit.com (Brian Bruns) Date: Tue Jan 4 14:51:33 2005 Subject: [Rt-devel] Bug in output of HTML code in ticket under History? Message-ID: <00e501c4f295$c2400180$05010e0a@2mbit.com> Hello all, Sorry if I am reposting a bug that may already have been fixed at some point, but I figured I'd mention it. I've just recently upgraded to 3.4.0rc1 from 3.2.2 (although this happens with a clean stock version of 3.4.0rc1 as well). In the ticket display, under the ticket history, where it shows the headers of the initial message, instead of showing links, it shows HTML code, due in part to RT outputting HTML character codes < instead of < for example. I can give examples, and access to the RT system in question for debugging if necessary. This machine runs Apache 1.3.33, and RT is running under mod_fastcgi. Once again, sorry if this post is a duplicate, and I'd be more then happy to help out however I can. -- Brian Bruns The Summit Open Source Development Group http://www.sosdg.org / http://www.ahbl.org From jesse at bestpractical.com Tue Jan 4 15:08:31 2005 From: jesse at bestpractical.com (Jesse Vincent) Date: Tue Jan 4 15:08:37 2005 Subject: [Rt-devel] Bug in output of HTML code in ticket under History? In-Reply-To: <00e501c4f295$c2400180$05010e0a@2mbit.com> References: <00e501c4f295$c2400180$05010e0a@2mbit.com> Message-ID: <20050104200831.GH24460@bestpractical.com> On Tue, Jan 04, 2005 at 02:44:06PM -0500, Brian Bruns wrote: > Hello all, > > Sorry if I am reposting a bug that may already have been fixed at some > point, but I figured I'd mention it. I've just recently upgraded to > 3.4.0rc1 from 3.2.2 (although this happens with a clean stock version > of 3.4.0rc1 as well). > > In the ticket display, under the ticket history, where it shows the > headers of the initial message, instead of showing links, it shows > HTML code, due in part to RT outputting HTML character codes < > instead of < for example. I can give examples, and access to the RT > system in question for debugging if necessary. Examples are useful. How can we reproduce this issue? > This machine runs Apache 1.3.33, and RT is running under mod_fastcgi. > > Once again, sorry if this post is a duplicate, and I'd be more then > happy to help out however I can. > -- > Brian Bruns > The Summit Open Source Development Group > http://www.sosdg.org / http://www.ahbl.org > > _______________________________________________ > Rt-devel mailing list > Rt-devel@lists.bestpractical.com > http://lists.bestpractical.com/cgi-bin/mailman/listinfo/rt-devel > -- From bruns at 2mbit.com Tue Jan 4 15:21:10 2005 From: bruns at 2mbit.com (Brian Bruns) Date: Tue Jan 4 15:23:03 2005 Subject: [Rt-devel] Bug in output of HTML code in ticket under History? References: <00e501c4f295$c2400180$05010e0a@2mbit.com> <20050104200831.GH24460@bestpractical.com> Message-ID: <010901c4f29a$effb3680$05010e0a@2mbit.com> On Tuesday, January 04, 2005 3:08 PM [EST], Jesse Vincent wrote: > Examples are useful. How can we reproduce this issue? Basically, its happening all the time on new tickets created. In the test cases I've done, I've basically sent an email to the rt-ahbl-request@sosdg.org address, which creates a new ticket in the AHBL Removal Requests Queue (If you do send an e-mail to this address, please let me know - the system will generate a password for you, and you can use the interface to look at what I mean for yourself if you need to) Thats about all I can say about reproducing the issue - I've attached a text file with the HTML output of what ends up happening. About all I can say at this point is the obvious. I wish I was better with perl and the various modules you can use, but I've got most of my minimal development knowledge in PHP. I am however, doing the best I can to try and trace where the problem might be coming from. Thanks again. -- Brian Bruns The Summit Open Source Development Group http://www.sosdg.org / http://www.ahbl.org -------------- next part -------------- #17: TEST
The Summit Open Source Development Group
Open Solutions For A Closed World / The Abusive Hosts Blocking List
    Skip Menu | Preferences | Logout
Logged in as bruns
RT for SOSDG
 
 
#17: TEST
Reply | Resolve | Open | Take | Comment
X  Ticket metadata  
X  The Basics  
Id: 17
Status: new
Left: 0 min
Priority: 1/5
Queue: AHBL Removal Requests

X  Custom Fields  
Assocated IP range(s):
  • (no value)

X  People  
Owner: Nobody
Requestors: bruns <bruns@2mbit.com>
Cc:
AdminCc:

X  Dates  
Created: Tue Jan 04 15:12:46 2005
Starts: Not set
Started: Not set
Last Contact: Not set
Due: Tue Jan 11 15:12:46 2005
Closed: Not set
Updated: Tue Jan 04 15:12:49 2005 by bruns

X  Links  
Depends on:
Depended on by:
Parents:
Children:
Refers to:
Referred to by:


X  History Display mode:[Brief headers] [Full headers]
#     Tue Jan 04 15:12:47 2005  bruns - Ticket created    [Reply] [Comment]
Subject: <A HREF="/RTIR/Tools/Lookup.html?ticket=17&type=handle&q=TEST">TEST</A><A HREF="/RTIR/Tools/Lookup.html?ticket=17&type=handle&q=X-">X-</A>Sa-Exim-Mail-From: <A HREF="/RTIR/Tools/Lookup.html?ticket=17&type=email&q=bruns@2mbit.com (2mbit.com)">bruns@2mbit.com (2mbit.com)</A> <A HREF="/RTIR/Tools/Lookup.html?ticket=17&type=host&q=2mbit.com">(2mbit.com)</A><A HREF="/RTIR/Tools/Lookup.html?ticket=17&type=handle&q=X-">X-</A>Msmail-Priority: Normal<A HREF="/RTIR/Tools/Lookup.html?ticket=17&type=handle&q=X-">X-</A>Priority: 3<A HREF="/RTIR/Tools/Lookup.html?ticket=17&type=handle&q=X-">X-</A>Sa-Exim-Scanned: No (on <A HREF="/RTIR/Tools/Lookup.html?ticket=17&type=host&q=mail.sosdg.org">mail.sosdg.org</A>); SAEximRunCond expanded to false
Date: Tue, 4 Jan 2005 15:12:38 -0500<A HREF="/RTIR/Tools/Lookup.html?ticket=17&type=handle&q=X-">X-</A>Mimeole: Produced By Microsoft MimeOLE <A HREF="/RTIR/Tools/Lookup.html?ticket=17&type=handle&q=V6">V6</A>.00.2800.1478
To: <<A HREF="/RTIR/Tools/Lookup.html?ticket=17&type=email&q=rt-ahbl-request@sosdg.org (sosdg.org)">rt-ahbl-request@sosdg.org (sosdg.org)</A> <A HREF="/RTIR/Tools/Lookup.html?ticket=17&type=host&q=sosdg.org">(sosdg.org)</A>>
From: "Brian Bruns" <<A HREF="/RTIR/Tools/Lookup.html?ticket=17&type=email&q=bruns@2mbit.com (2mbit.com)">bruns@2mbit.com (2mbit.com)</A> <A HREF="/RTIR/Tools/Lookup.html?ticket=17&type=host&q=2mbit.com">(2mbit.com)</A>><A HREF="/RTIR/Tools/Lookup.html?ticket=17&type=handle&q=X-RT-">X-RT-</A>Original-Encoding: iso-8859-1
Download (untitled)
text/plain 104b

--
Brian Bruns
The Summit Open Source Development Group
http://www.sosdg.org (www.sosdg.org) / http://www.ahbl.org (www.ahbl.org)
#     Tue Jan 04 15:12:48 2005  RT_System - Outgoing email recorded    [Show] 
»|« RT 3.4.0rc1 Copyright 1996-2004 Best Practical Solutions, LLC.
Time to display: 1.36168
From jesse at bestpractical.com Tue Jan 4 15:30:43 2005 From: jesse at bestpractical.com (Jesse Vincent) Date: Tue Jan 4 15:30:49 2005 Subject: [Rt-devel] Bug in output of HTML code in ticket under History? In-Reply-To: <010901c4f29a$effb3680$05010e0a@2mbit.com> References: <00e501c4f295$c2400180$05010e0a@2mbit.com> <20050104200831.GH24460@bestpractical.com> <010901c4f29a$effb3680$05010e0a@2mbit.com> Message-ID: <20050104203042.GK24460@bestpractical.com> On Tue, Jan 04, 2005 at 03:21:10PM -0500, Brian Bruns wrote: > On Tuesday, January 04, 2005 3:08 PM [EST], Jesse Vincent wrote: > > Examples are useful. How can we reproduce this issue? > > Basically, its happening all the time on new tickets created. In the > test cases I've done, I've basically sent an email to the > rt-ahbl-request@sosdg.org address, which creates a new ticket in the > AHBL Removal Requests Queue (If you do send an e-mail to this address, > please let me know - the system will generate a password for you, and > you can use the interface to look at what I mean for yourself if you > need to) Oh. You're attempting to use RTIR 1.0.x with RT 3.4. RT 3.2 and newer are not compatible with with RTIR 1.0.x. Try the 1.1 RC instead? From bruns at 2mbit.com Tue Jan 4 16:08:52 2005 From: bruns at 2mbit.com (Brian Bruns) Date: Tue Jan 4 16:10:48 2005 Subject: [Rt-devel] Bug in output of HTML code in ticket under History? References: <00e501c4f295$c2400180$05010e0a@2mbit.com> <20050104200831.GH24460@bestpractical.com> <010901c4f29a$effb3680$05010e0a@2mbit.com> <20050104203042.GK24460@bestpractical.com> Message-ID: <012d01c4f2a1$9a35a580$05010e0a@2mbit.com> On Tuesday, January 04, 2005 3:30 PM [EST], Jesse Vincent wrote: > > Oh. You're attempting to use RTIR 1.0.x with RT 3.4. RT 3.2 and > newer are not compatible with with RTIR 1.0.x. Try the 1.1 RC > instead? Did I mention how much of an idiot I feel like right now? :) I'm sorry for bugging you guys about this. Let me know if theres anything I can do to make it up to you. -- Brian Bruns The Summit Open Source Development Group http://www.sosdg.org / http://www.ahbl.org From jesse at bestpractical.com Tue Jan 4 16:20:36 2005 From: jesse at bestpractical.com (Jesse Vincent) Date: Tue Jan 4 16:20:40 2005 Subject: [Rt-devel] Bug in output of HTML code in ticket under History? In-Reply-To: <012d01c4f2a1$9a35a580$05010e0a@2mbit.com> References: <00e501c4f295$c2400180$05010e0a@2mbit.com> <20050104200831.GH24460@bestpractical.com> <010901c4f29a$effb3680$05010e0a@2mbit.com> <20050104203042.GK24460@bestpractical.com> <012d01c4f2a1$9a35a580$05010e0a@2mbit.com> Message-ID: <20050104212036.GN24460@bestpractical.com> On Tue, Jan 04, 2005 at 04:08:52PM -0500, Brian Bruns wrote: > On Tuesday, January 04, 2005 3:30 PM [EST], Jesse Vincent wrote: > > > > > Oh. You're attempting to use RTIR 1.0.x with RT 3.4. RT 3.2 and > > newer are not compatible with with RTIR 1.0.x. Try the 1.1 RC > > instead? > > Did I mention how much of an idiot I feel like right now? :) > > I'm sorry for bugging you guys about this. Let me know if theres > anything I can do to make it up to you. The best way to help out when you feel like you've been a support burden is to help reduce the burden for others. The next time you see someone ask a question that you know the answer to, put together a Q&A page on wiki.bestpractical.com. Then point the user to that wiki page in a reply to the list. Thanks! Jesse > -- > Brian Bruns > The Summit Open Source Development Group > http://www.sosdg.org / http://www.ahbl.org > -- From matthew.watson at staff.netspace.net.au Tue Jan 4 21:59:59 2005 From: matthew.watson at staff.netspace.net.au (Matthew Watson) Date: Tue Jan 4 22:14:34 2005 Subject: [Rt-devel] wiki pages invaded Message-ID: <56F211C5E3F24F47B103EA1B253822BE243F19@vic-cr-ex1.staff.netspace.net.au> Heya. Just a heads up, looking around the wiki site, looks like it has been invaded by someone with too much time and they have left spam all over the place. Regards -- Matthew Watson Netspace Online Systems (03) 98110010 -------------- next part -------------- An HTML attachment was scrubbed... URL: http://bestpractical.com/pipermail/rt-devel/attachments/20050105/f7cc261f/attachment.htm From ruediger.riediger at sun.com Thu Jan 6 03:55:20 2005 From: ruediger.riediger at sun.com (Ruediger Riediger) Date: Thu Jan 6 03:57:31 2005 Subject: [Rt-devel] Last Transaction anchor placed wrong In-Reply-To: <20041215191002.GU22556@bestpractical.com> References: <41C08520.1080206@sun.com> <20041215191002.GU22556@bestpractical.com> Message-ID: <41DCFCF8.40004@sun.com> Jesse Vincent wrote: > Hm. That looks like a perfectly valid issue. (I agree that it's broken) > but it's always worked just fine for me. What browser is it biting you > in? Same issue: Mozilla/5.0 (X11; U; SunOS sun4u; en-US; rv:1.7.5) Gecko/20041110 Firefox/1.0 Best regards, Ruediger Riediger -- Dr. Ruediger Riediger Sun Microsystems GmbH NSG - SunCERT Komturstr. 18a mailto:Ruediger.Riediger@Sun.com D-12099 Berlin ------------------------------------------------------------------------ NOTICE: This email message is for the sole use of the intended recipient(s) and may contain confidential and privileged information. Any unauthorized review, use, disclosure or distribution is prohibited. If you are not the intended recipient, please contact the sender by reply email and destroy all copies of the original message. ------------------------------------------------------------------------ PGP 2048RSA/0x2C5020E9 964C E189 0FF0 8882 2BAB 65E2 6912 1FF2 ------------------------------------------------------------------------ From jesse at bestpractical.com Thu Jan 6 16:29:25 2005 From: jesse at bestpractical.com (Jesse Vincent) Date: Thu Jan 6 16:29:32 2005 Subject: [Rt-devel] RT 3.4.0rc2 Message-ID: <20050106212925.GX24460@bestpractical.com> RT 3.4.0rc2 fixes a number of small bugs in RT 3.4.0rc1. Most notably, the oracle upgrade procedure now works correctly. DBD::Oracle 1.16 support is still not quite there, though with DBIx::SearchBuilder 1.18, we're down to a single (very important) failing test. Jesse ------------------------------------------------------------------------ r2044 | jesse | 2005-01-06 16:00:57 -0500 (Thu, 06 Jan 2005) | 3 lines Changed paths: M /rt/branches/3.4-RELEASE M /rt/branches/3.4-RELEASE/lib/t/regression/04send_email.t M /rt/branches/3.4-RELEASE/lib/t/regression/06mailgateway.t r2543@hualien: jesse | 2005-01-06T07:08:12.346569Z no longer hardcode paths in tests ------------------------------------------------------------------------ r2041 | jesse | 2005-01-05 00:34:02 -0500 (Wed, 05 Jan 2005) | 3 lines Changed paths: M /rt/branches/3.4-RELEASE M /rt/branches/3.4-RELEASE/lib/RT/Record.pm M /rt/branches/3.4-RELEASE/lib/t/regression/03basic_web.t r2539@hualien: jesse | 2005-01-05T05:31:36.967112Z Better handling of Content that comes out of the database as utf8 (Such as with DBD::Oracle 1.16) ------------------------------------------------------------------------ r2040 | jesse | 2005-01-04 00:18:06 -0500 (Tue, 04 Jan 2005) | 3 lines Changed paths: M /rt/branches/3.4-RELEASE M /rt/branches/3.4-RELEASE/lib/RT/Ticket_Overlay.pm r2530@hualien: jesse | 2005-01-04T05:15:35.757280Z a couple fixes for better handling ticket links on merge ------------------------------------------------------------------------ r2039 | jesse | 2005-01-04 00:17:52 -0500 (Tue, 04 Jan 2005) | 4 lines Changed paths: M /rt/branches/3.4-RELEASE A /rt/branches/3.4-RELEASE/lib/t/regression/11-template-insert.t r2528@hualien: jesse | 2005-01-03T19:54:59.819442Z Added a test for a regression tickled by DBD::Oracle 1.16 ------------------------------------------------------------------------ r2038 | jesse | 2005-01-04 00:17:42 -0500 (Tue, 04 Jan 2005) | 5 lines Changed paths: M /rt/branches/3.4-RELEASE M /rt/branches/3.4-RELEASE/etc/upgrade/3.3.0/schema.mysql r2527@hualien: jesse | 2005-01-03T19:52:28.164258Z Updated mysql schema upgrades to cope with ancient RT 3 instances which didn't have indexes we expect ------------------------------------------------------------------------ r2037 | jesse | 2005-01-04 00:17:29 -0500 (Tue, 04 Jan 2005) | 4 lines Changed paths: M /rt/branches/3.4-RELEASE M /rt/branches/3.4-RELEASE/etc/upgrade/3.3.0/schema.mysql r2526@hualien: jesse | 2005-01-03T19:21:48.180206Z Workaround for missing mysql indexes ------------------------------------------------------------------------ r2036 | jesse | 2005-01-03 14:14:54 -0500 (Mon, 03 Jan 2005) | 5 lines Changed paths: M /rt/branches/3.4-RELEASE A /rt/branches/3.4-RELEASE/etc/upgrade/3.3.0/schema.Oracle r2524@hualien: jesse | 2005-01-03T19:12:32.074066Z Oracle schema upgrades ------------------------------------------------------------------------ r2034 | jesse | 2004-12-29 15:41:21 -0500 (Wed, 29 Dec 2004) | 5 lines Changed paths: M /rt/branches/3.4-RELEASE M /rt/branches/3.4-RELEASE/html/Admin/CustomFields/GroupRights.html M /rt/branches/3.4-RELEASE/html/Admin/CustomFields/Modify.html M /rt/branches/3.4-RELEASE/html/Admin/CustomFields/Objects.html M /rt/branches/3.4-RELEASE/html/Admin/CustomFields/UserRights.html M /rt/branches/3.4-RELEASE/html/Admin/CustomFields/index.html M /rt/branches/3.4-RELEASE/html/Admin/Elements/CustomFieldTabs r2520@hualien: jesse | 2004-12-29T20:39:43.707258Z Custom field tab hilighting fixes. ------------------------------------------------------------------------ r2033 | jesse | 2004-12-29 14:45:54 -0500 (Wed, 29 Dec 2004) | 5 lines Changed paths: M /rt/branches/3.4-RELEASE M /rt/branches/3.4-RELEASE/lib/RT/Handle.pm r2516@hualien: jesse | 2004-12-29T19:44:09.437463Z Slightly updated encoding support for oracle ------------------------------------------------------------------------ r2032 | jesse | 2004-12-29 14:45:43 -0500 (Wed, 29 Dec 2004) | 4 lines Changed paths: M /rt/branches/3.4-RELEASE M /rt/branches/3.4-RELEASE/UPGRADING r2507@hualien: jesse | 2004-12-23T22:31:20.362197Z Note about "OwnTickets" for Superusers ------------------------------------------------------------------------ r2030 | robert | 2004-12-28 01:37:51 -0500 (Tue, 28 Dec 2004) | 3 lines Changed paths: M /rt/branches/3.4-RELEASE M /rt/branches/3.4-RELEASE/configure.ac r2033@dog: rspier | 2004-12-27 22:36:08 -0800 RT 6331: configure spitting chmod errors ------------------------------------------------------------------------ r2029 | robert | 2004-12-28 01:37:47 -0500 (Tue, 28 Dec 2004) | 3 lines Changed paths: M /rt/branches/3.4-RELEASE M /rt/branches/3.4-RELEASE/html/NoAuth/webrt.css r2032@dog: rspier | 2004-12-27 22:24:34 -0800 fix css errors ------------------------------------------------------------------------ r2028 | robert | 2004-12-28 01:37:37 -0500 (Tue, 28 Dec 2004) | 4 lines Changed paths: M /rt/branches/3.4-RELEASE M /rt/branches/3.4-RELEASE/html/Elements/Header r2031@dog: rspier | 2004-12-27 21:46:34 -0800 Only show the refresh tag if the refresh time is > 0. (Mozilla will perform a refresh=1 for refresh=-1) ------------------------------------------------------------------------ r2027 | robert | 2004-12-28 01:37:32 -0500 (Tue, 28 Dec 2004) | 3 lines Changed paths: M /rt/branches/3.4-RELEASE M /rt/branches/3.4-RELEASE/configure.ac r2030@dog: rspier | 2004-12-27 21:05:44 -0800 Whitespace tweak: use tabs consistently ------------------------------------------------------------------------ r2025 | jesse | 2004-12-22 02:06:35 -0500 (Wed, 22 Dec 2004) | 5 lines Changed paths: M /rt/branches/3.4-RELEASE M /rt/branches/3.4-RELEASE/etc/schema.Oracle r2504@hualien: jesse | 2004-12-22T07:05:58.453205Z Oracle schema updates ------------------------------------------------------------------------ r2024 | jesse | 2004-12-21 17:11:39 -0500 (Tue, 21 Dec 2004) | 7 lines Changed paths: M /rt/branches/3.4-RELEASE M /rt/branches/3.4-RELEASE/html/Ticket/Elements/ShowHistory M /rt/branches/3.4-RELEASE/lib/RT/Ticket_Overlay.pm M /rt/branches/3.4-RELEASE/lib/RT/Transactions_Overlay.pm r2500@hualien: jesse | 2004-12-21T22:06:43.378557Z Refactored RT's logic that limits sets of transactions to tickets to make it easier to reuse. Fixed bugs in searches for transactions related to multiple tickets. ------------------------------------------------------------------------ r2021 | jesse | 2004-12-21 00:32:15 -0500 (Tue, 21 Dec 2004) | 5 lines Changed paths: M /rt/branches/3.4-RELEASE A /rt/branches/3.4-RELEASE/lib/t/regression/10merge.t r2498@hualien: jesse | 2004-12-21T05:27:21.459146Z Added a new regression test to validate some possibly bogus behaviour about merging tickets ------------------------------------------------------------------------ r2019 | jesse | 2004-12-20 21:44:53 -0500 (Mon, 20 Dec 2004) | 6 lines Changed paths: M /rt/branches/3.4-RELEASE M /rt/branches/3.4-RELEASE/lib/RT/Tickets_Overlay.pm r2494@hualien: jesse | 2004-12-21T02:40:37.522278Z Added back a 'package' declaration to Tickets_Overlay to enable automatic reload. ------------------------------------------------------------------------ r2018 | jesse | 2004-12-20 18:53:18 -0500 (Mon, 20 Dec 2004) | 4 lines Changed paths: M /rt/branches/3.4-RELEASE M /rt/branches/3.4-RELEASE/bin/mason_handler.fcgi.in r2492@hualien: jesse | 2004-12-20T23:49:03.808597Z The FastCGI handler was missing a small bit of code to enable the devel mode "Module::Reload" feature. ------------------------------------------------------------------------ r2017 | jesse | 2004-12-20 18:33:50 -0500 (Mon, 20 Dec 2004) | 5 lines Changed paths: M /rt/branches/3.4-RELEASE M /rt/branches/3.4-RELEASE/lib/t/regression/08web_cf_access.t r2489@hualien: jesse | 2004-12-20T23:27:06.068915Z Inhibit a warning in 08_web_Cf_access.t From luiz at eaprender.com.br Fri Jan 7 06:07:58 2005 From: luiz at eaprender.com.br (Luiz Henrique Ozaki) Date: Fri Jan 7 06:15:21 2005 Subject: [Rt-devel] rt-mailgate problems Message-ID: <41DE6D8E.9010600@eaprender.com.br> Im using pgsql alias map with postfix and I doesnt execute that rt-mailgate command when a mail arrives to that alias. How can I make it working using pgsql alias map ? Here goes the sql data. address | goto | domain | created | modified | active ---------------------+-----------------------------------------------------------------------------------------------+----------+-------------------------------+-------------------------------+-------- rt@ibep.net | |/usr/local/rt3/bin/rt-mailgate --queue general --action correspond --url http://localhost/ | ibep.net | 2005-01-04 11:42:57.176243-02 | 2005-01-04 11:42:57.176243-02 | t rt-comment@ibep.net | "|/usr/local/rt3/bin/rt-mailgate --queue general --action correspond --url http://localhost/" | ibep.net | 2005-01-04 11:42:57.176243-02 | 2005-01-04 11:42:57.176243-02 | t Regards, -- []'s Luiz Henrique Ozaki From dkrt at navahonetworks.com Fri Jan 7 13:08:07 2005 From: dkrt at navahonetworks.com (David Kerry) Date: Fri Jan 7 13:29:44 2005 Subject: [Rt-devel] 3.4.0rc1 slow Message-ID: <20050107180807.GA8887@navahonetworks.com> Hello, I've been spending some time with 3.4.0rc1 lately with some of our existing data and have found some issues related to slow or excessive database access. Setup: Postgres 7.4.6 RT 3.4.0rc1 SearchBuilder 1.16 Apache 2+mod_perl2 (Debian sarge) Hardware: dual 3Ghz Xeon server Both rt and database are on the same machine. I have 2 issues: 1) The query builder page (when you hit the 'tickets' link) is exceptionally slow to initially come up. About 10 seconds just to bring up the empty form. Is this normal? I've looked at the postgres logs, and it seems to do many many iterations of this type of query: LOG: duration: 1983.993 ms statement: SELECT DISTINCT main.* FROM (Groups main LEFT JOIN Principals Principals_2 ON ( main.id = Principals_2.id)), ACL ACL_1 WHERE ((ACL_1.RightName = 'OwnTicket')) AND ((Principals_2.Disabled = '0')) AND ( ( ACL_1.PrincipalId = main.id AND ACL_1.PrincipalType = 'Group' AND ( main.Domain = 'SystemInternal' OR main.Domain = 'UserDefined' OR main.Domain = 'ACLEquivalence')) OR ( ( (main.Domain = 'RT::Queue-Role' AND main.Instance = 102) ) AND main.Type = ACL_1.PrincipalType) ) AND (ACL_1.ObjectType = 'RT::System' OR (ACL_1.ObjectType = 'RT::Queue' AND ACL_1.ObjectId = 102) ) ORDER BY main.Name ASC At the time, there were about 50k+ users and 20k+ groups in the database. 2) The method used in the Postgres version of SearchBuilder/Handle/Pg.pm for insert() is suboptimal. I've turned on postgres logging to dump any queries taking > .5 seconds and the query in this method comes up continually: ie: postgres[18131]: [6610-1] LOG: duration: 881.154 ms statement: SELECT id FROM Attachments WHERE oid = '3609306' From doing a bit of reading in the postgres docs, the oid should not be used for this type of thing - they can/will roll, which can lead to duplicates too. The same query, selecting on 'id' from the Attachments table is virtually instantaneous. I would highly recommend taking the approach used in the Oracle version of this class and select the next id from a sequence and insert using that instead of depending on the database to auto-generate an id and then having to figure out what id it generated. Alternately - use guids everywhere instead of sequences/ids - easy to generate, avoids duplicate key issues on imports and has many other benefits. This becomes a serious issue when you throw database replication into the mix too. -- David Kerry From luiz at eaprender.com.br Fri Jan 7 13:40:37 2005 From: luiz at eaprender.com.br (Luiz Henrique Ozaki) Date: Fri Jan 7 13:42:55 2005 Subject: [Rt-devel] Re: [rt-users] Mail Gateway problems In-Reply-To: References: <41DD81E7.4030206@eaprender.com.br> Message-ID: <41DED7A5.1000403@eaprender.com.br> Andy Harrison wrote: >>postfix=> >>select * from alias WHERE address='rt'; >> address | goto >> | domain | created | >> modified | active >>---------+---------------------------------------------------------------------------------------------+----------+-------------------------------+-------------------------------+-------- >> rt | |/usr/local/rt3/bin/rt-mailgate --queue general --action >>correspond --url http://localhost/ | ibep.net | 2005-01-04 >>11:42:57.176243-02 | 2005-01-04 11:42:57.176243-02 | t >> (1 row) >> >> >> > >Try changing the queue to $DEFAULT. Here is my mailgate command: > >|/usr/local/rt3/bin/rt-mailgate --queue $DEFAULT --action correspond >--url http://localhost > > > > Here goes the solution ! http://wiki.bestpractical.com/index.cgi?SQLalias -- []'s Luiz Henrique Ozaki -------------- next part -------------- An HTML attachment was scrubbed... URL: http://bestpractical.com/pipermail/rt-devel/attachments/20050107/e05288f9/attachment.htm From jesse at bestpractical.com Fri Jan 7 17:59:08 2005 From: jesse at bestpractical.com (Jesse Vincent) Date: Fri Jan 7 17:59:13 2005 Subject: [Rt-devel] 3.4.0rc1 slow In-Reply-To: <20050107180807.GA8887@navahonetworks.com> References: <20050107180807.GA8887@navahonetworks.com> Message-ID: <20050107225908.GX24460@bestpractical.com> > I've looked at the postgres logs, and it seems to do many many iterations > of this type of query: > > LOG: duration: 1983.993 ms statement: > SELECT DISTINCT main.* FROM (Groups main LEFT JOIN Principals Principals_2 ON ( main.id = Principals_2.id)), > ACL ACL_1 WHERE ((ACL_1.RightName = 'OwnTicket')) AND ((Principals_2.Disabled = '0')) AND ( ( ACL_1.PrincipalId = main.id AND > ACL_1.PrincipalType = 'Group' AND ( main.Domain = 'SystemInternal' OR main.Domain = 'UserDefined' OR main.Domain = > 'ACLEquivalence')) OR ( ( (main.Domain = 'RT::Queue-Role' AND main.Instance = 102) ) AND main.Type = ACL_1.PrincipalType) ) AND > (ACL_1.ObjectType = 'RT::System' OR (ACL_1.ObjectType = 'RT::Queue' AND ACL_1.ObjectId = 102) ) ORDER BY main.Name ASC > > At the time, there were about 50k+ users and 20k+ groups in the > database. I'd love it if you could dig a bit more into what's killing that query. Pull off the distinct and see if we're doing something stupid and generating a huge cartesian porduct? > > > 2) The method used in the Postgres version of SearchBuilder/Handle/Pg.pm > for insert() is suboptimal. I've turned on postgres logging to dump > any queries taking > .5 seconds and the query in this method comes up > continually: Interesting. At the time we did that (years ago now), that was the recommended way to do unique ids on postgres. I'd love to see a patch for improved behaviour. > ie: > > postgres[18131]: [6610-1] LOG: duration: 881.154 ms statement: SELECT id FROM Attachments WHERE oid = '3609306' > > From doing a bit of reading in the postgres docs, the oid should not > be used for this type of thing - they can/will roll, which can lead > to duplicates too. But not instantly. And the only time we're doing that is immediately after the insert. > The same query, selecting on 'id' from the Attachments table is > virtually instantaneous. > From ajs at crankycanuck.ca Fri Jan 7 18:13:34 2005 From: ajs at crankycanuck.ca (Andrew Sullivan) Date: Fri Jan 7 18:22:50 2005 Subject: [Rt-devel] 3.4.0rc1 slow In-Reply-To: <20050107225908.GX24460@bestpractical.com> References: <20050107180807.GA8887@navahonetworks.com> <20050107225908.GX24460@bestpractical.com> Message-ID: <20050107231334.GF14466@phlogiston.dyndns.org> On Fri, Jan 07, 2005 at 05:59:08PM -0500, Jesse Vincent wrote: > > From doing a bit of reading in the postgres docs, the oid should not > > be used for this type of thing - they can/will roll, which can lead > > to duplicates too. > > But not instantly. And the only time we're doing that is immediately > after the insert. More importantly, Postgres won't have oids on most tables by default in the near future. So it won't even work unless your DBA knows to flip the WITH OIDS option. A -- Andrew Sullivan | ajs@crankycanuck.ca A certain description of men are for getting out of debt, yet are against all taxes for raising money to pay it off. --Alexander Hamilton From jesse at bestpractical.com Fri Jan 7 22:43:21 2005 From: jesse at bestpractical.com (Jesse Vincent) Date: Fri Jan 7 22:43:28 2005 Subject: [Rt-devel] RT 3.4.0rc3 Message-ID: <20050108034321.GY24460@bestpractical.com> I'm somewhat ashamed to announce the release of RT 3.4.0rc3. This release features a retraction of comming 1976, as well as new tests to ensure that the particular errors caused by that issue never happen again. (Primarily ticket watcher and custom field searching were broken.) Jesse -- From jesse at bestpractical.com Fri Jan 7 23:10:19 2005 From: jesse at bestpractical.com (Jesse Vincent) Date: Fri Jan 7 23:10:23 2005 Subject: [Rt-devel] 3.4.0rc1 slow In-Reply-To: <20050107231334.GF14466@phlogiston.dyndns.org> References: <20050107180807.GA8887@navahonetworks.com> <20050107225908.GX24460@bestpractical.com> <20050107231334.GF14466@phlogiston.dyndns.org> Message-ID: <20050108041018.GZ24460@bestpractical.com> > > More importantly, Postgres won't have oids on most tables by default > in the near future. So it won't even work unless your DBA knows to > flip the WITH OIDS option. It was only a year or two that I was seeing people widely recommending the strategy SB is using to get insert ids. There are going to be lots of sad webapp developers out there :/ But at least I can fix it easily. > > A > > > -- > Andrew Sullivan | ajs@crankycanuck.ca > A certain description of men are for getting out of debt, yet are > against all taxes for raising money to pay it off. > --Alexander Hamilton > _______________________________________________ > Rt-devel mailing list > Rt-devel@lists.bestpractical.com > http://lists.bestpractical.com/cgi-bin/mailman/listinfo/rt-devel > -- From dkrt at navahonetworks.com Sat Jan 8 00:28:49 2005 From: dkrt at navahonetworks.com (David Kerry) Date: Sat Jan 8 00:31:02 2005 Subject: [Rt-devel] 3.4.0rc1 slow In-Reply-To: <20050107225908.GX24460@bestpractical.com> References: <20050107180807.GA8887@navahonetworks.com> <20050107225908.GX24460@bestpractical.com> Message-ID: <20050108052849.GA14618@navahonetworks.com> On Fri, Jan 07, 2005 at 05:59:08PM -0500, Jesse Vincent wrote: > > > I've looked at the postgres logs, and it seems to do many many iterations > > of this type of query: > > > > LOG: duration: 1983.993 ms statement: > > SELECT DISTINCT main.* FROM (Groups main LEFT JOIN Principals Principals_2 ON ( main.id = Principals_2.id)), > > ACL ACL_1 WHERE ((ACL_1.RightName = 'OwnTicket')) AND ((Principals_2.Disabled = '0')) AND ( ( ACL_1.PrincipalId = main.id AND > > ACL_1.PrincipalType = 'Group' AND ( main.Domain = 'SystemInternal' OR main.Domain = 'UserDefined' OR main.Domain = > > 'ACLEquivalence')) OR ( ( (main.Domain = 'RT::Queue-Role' AND main.Instance = 102) ) AND main.Type = ACL_1.PrincipalType) ) AND > > (ACL_1.ObjectType = 'RT::System' OR (ACL_1.ObjectType = 'RT::Queue' AND ACL_1.ObjectId = 102) ) ORDER BY main.Name ASC > > > > At the time, there were about 50k+ users and 20k+ groups in the > > database. > > I'd love it if you could dig a bit more into what's killing that query. > Pull off the distinct and see if we're doing something stupid and > generating a huge cartesian porduct? Ok - my mistake - I hadn't done an analyze on the tables/indexes when I ran that (was taking 30+ seconds without the analyze to bring up the query page). After the analyze, that same query drops down to ~500ms. Still fairly long given the hardware I'm running on, however. I've played with the query a bit, and removing the distinct doesn't really change things much. The result is 8 rows returned instead of 7. Removing the distinct and order by results in the same cost by the optimizer (according to 'explain', anyways). I also did some more investigation and there are other things amiss I think. Simply bringing up the tickets->query builder page generates over 129 sql queries ?! 19 of those queries took between 400 and 600ms apiece. That's an awful lot of data trolling to bring up one screen. Is that right, or is my setup misbehaving/borked somehow? I have saved the entire sql trace in case anyone is interested in looking at it in detail. Everything else so far seems to be reasonable snappy except for this one screen, which is unfortunate, because my users will be running into that particular one continuously. > > 2) The method used in the Postgres version of SearchBuilder/Handle/Pg.pm > > for insert() is suboptimal. I've turned on postgres logging to dump > > any queries taking > .5 seconds and the query in this method comes up > > continually: > > Interesting. At the time we did that (years ago now), that was the > recommended way to do unique ids on postgres. I'd love to see a patch > for improved behaviour. I've got a patch worked up for this - basically I just copied the oracle way (which should be many many times faster, but I still have to test it). This should give improved insert performance (which has been an issue for us in the past as well under particularly high incoming mail volume). -- David Kerry From jesse at bestpractical.com Sat Jan 8 01:21:57 2005 From: jesse at bestpractical.com (Jesse Vincent) Date: Sat Jan 8 01:22:04 2005 Subject: [Rt-devel] 3.4.0rc1 slow In-Reply-To: <20050108052849.GA14618@navahonetworks.com> References: <20050107180807.GA8887@navahonetworks.com> <20050107225908.GX24460@bestpractical.com> <20050108052849.GA14618@navahonetworks.com> Message-ID: <20050108062157.GC24460@bestpractical.com> > > Simply bringing up the tickets->query builder page generates over 129 sql > queries ?! 19 of those queries took between 400 and 600ms apiece. That's > an awful lot of data trolling to bring up one screen. Is that right, or > is my setup misbehaving/borked somehow? I have saved the entire sql trace > in case anyone is interested in looking at it in detail. > My first guess is that it's mostly in the dropdowns for picking ticket owners by name or group membership. Using MasonX::Profiler, you might be able to see which components are causing the slowdowns, which will help to find out how to fix em. Was 3.2 similarly slow? > Everything else so far seems to be reasonable snappy except for this one > screen, which is unfortunate, because my users will be running into that > particular one continuously. From rt at chaka.net Sat Jan 8 01:04:51 2005 From: rt at chaka.net (Todd Chapman) Date: Sat Jan 8 01:46:19 2005 Subject: [Rt-devel] 3.4.0rc1 slow In-Reply-To: <20050108062157.GC24460@bestpractical.com> References: <20050107180807.GA8887@navahonetworks.com> <20050107225908.GX24460@bestpractical.com> <20050108052849.GA14618@navahonetworks.com> <20050108062157.GC24460@bestpractical.com> Message-ID: <20050108060451.GB27116@chaka.net> I have found that screen to be particularly slow in 3.2 -Todd On Sat, Jan 08, 2005 at 01:21:57AM -0500, Jesse Vincent wrote: > > > > > Simply bringing up the tickets->query builder page generates over 129 sql > > queries ?! 19 of those queries took between 400 and 600ms apiece. That's > > an awful lot of data trolling to bring up one screen. Is that right, or > > is my setup misbehaving/borked somehow? I have saved the entire sql trace > > in case anyone is interested in looking at it in detail. > > > > My first guess is that it's mostly in the dropdowns for picking ticket > owners by name or group membership. Using MasonX::Profiler, you might > be able to see which components are causing the slowdowns, which will > help to find out how to fix em. Was 3.2 similarly slow? > > > Everything else so far seems to be reasonable snappy except for this one > > screen, which is unfortunate, because my users will be running into that > > particular one continuously. > _______________________________________________ > Rt-devel mailing list > Rt-devel@lists.bestpractical.com > http://lists.bestpractical.com/cgi-bin/mailman/listinfo/rt-devel From ajs at crankycanuck.ca Sat Jan 8 10:54:23 2005 From: ajs at crankycanuck.ca (Andrew Sullivan) Date: Sat Jan 8 10:57:00 2005 Subject: [Rt-devel] 3.4.0rc1 slow In-Reply-To: <20050108041018.GZ24460@bestpractical.com> References: <20050107180807.GA8887@navahonetworks.com> <20050107225908.GX24460@bestpractical.com> <20050107231334.GF14466@phlogiston.dyndns.org> <20050108041018.GZ24460@bestpractical.com> Message-ID: <20050108155423.GA16063@phlogiston.dyndns.org> On Fri, Jan 07, 2005 at 11:10:19PM -0500, Jesse Vincent wrote: > > > > > More importantly, Postgres won't have oids on most tables by default > > in the near future. So it won't even work unless your DBA knows to > > flip the WITH OIDS option. > > It was only a year or two that I was seeing people widely recommending > the strategy SB is using to get insert ids. There are going to be lots > of sad webapp developers out there :/ But at least I can fix it easily. Yike. That's unfortunate, because it was always a bad idea, although not as potentially bad as it is now (ok, not quite always. Originally the oids were regarded as a feature, if you harken back to the days when it really was Postgres rather than PostgreSQL). But I certainly remember being cautioned by some against using them in pre-7.0 versions, and that's several years ago. That doesn't mean plenty of people weren't advocates of them as well. A -- Andrew Sullivan | ajs@crankycanuck.ca Information security isn't a technological problem. It's an economics problem. --Bruce Schneier From ajs at crankycanuck.ca Sat Jan 8 10:57:17 2005 From: ajs at crankycanuck.ca (Andrew Sullivan) Date: Sat Jan 8 10:59:55 2005 Subject: [Rt-devel] 3.4.0rc1 slow In-Reply-To: <20050108052849.GA14618@navahonetworks.com> References: <20050107180807.GA8887@navahonetworks.com> <20050107225908.GX24460@bestpractical.com> <20050108052849.GA14618@navahonetworks.com> Message-ID: <20050108155716.GB16063@phlogiston.dyndns.org> On Sat, Jan 08, 2005 at 12:28:49AM -0500, David Kerry wrote: > Ok - my mistake - I hadn't done an analyze on the tables/indexes when I ran > that (was taking 30+ seconds without the analyze to bring up the query page). I did wonder about this area, because it's something I know was recently improved. But this one. . . > Simply bringing up the tickets->query builder page generates over 129 sql > queries ?! 19 of those queries took between 400 and 600ms apiece. That's . . . is, according to one of the DBAs in my group at work, primarily due to the way the ACLs are processed. So, more objects for permissions entails more queries, and the delay expands linearly accorfdingly. We are also finding that screen to make the system extremely hard to use. A -- Andrew Sullivan | ajs@crankycanuck.ca This work was visionary and imaginative, and goes to show that visionary and imaginative work need not end up well. --Dennis Ritchie From jesse at bestpractical.com Sat Jan 8 16:01:56 2005 From: jesse at bestpractical.com (Jesse Vincent) Date: Sat Jan 8 16:02:01 2005 Subject: [Rt-devel] 3.4.0rc1 slow In-Reply-To: <20050108155716.GB16063@phlogiston.dyndns.org> References: <20050107180807.GA8887@navahonetworks.com> <20050107225908.GX24460@bestpractical.com> <20050108052849.GA14618@navahonetworks.com> <20050108155716.GB16063@phlogiston.dyndns.org> Message-ID: <20050108210156.GE24460@bestpractical.com> On Sat, Jan 08, 2005 at 10:57:17AM -0500, Andrew Sullivan wrote: > On Sat, Jan 08, 2005 at 12:28:49AM -0500, David Kerry wrote: > > Ok - my mistake - I hadn't done an analyze on the tables/indexes when I ran > > that (was taking 30+ seconds without the analyze to bring up the query page). > > I did wonder about this area, because it's something I know was > recently improved. But this one. . . > > > Simply bringing up the tickets->query builder page generates over 129 sql > > queries ?! 19 of those queries took between 400 and 600ms apiece. That's > > . . . is, according to one of the DBAs in my group at work, primarily > due to the way the ACLs are processed. So, more objects for > permissions entails more queries, and the delay expands linearly > accorfdingly. We are also finding that screen to make the system > extremely hard to use. I'd love to see suggestions about how we can improve the search UI without completely breaking it for users of downlevel browsers. (Which means that pretty javascript hacks and popup windows are pretty much right out.) Would people be satisfied if the pick lists for users and groups were transformed into text entry widgets? > > A > > -- > Andrew Sullivan | ajs@crankycanuck.ca > This work was visionary and imaginative, and goes to show that visionary > and imaginative work need not end up well. > --Dennis Ritchie > _______________________________________________ > Rt-devel mailing list > Rt-devel@lists.bestpractical.com > http://lists.bestpractical.com/cgi-bin/mailman/listinfo/rt-devel > -- From ajs at crankycanuck.ca Sat Jan 8 16:52:11 2005 From: ajs at crankycanuck.ca (Andrew Sullivan) Date: Sat Jan 8 16:54:49 2005 Subject: [Rt-devel] 3.4.0rc1 slow In-Reply-To: <20050108210156.GE24460@bestpractical.com> References: <20050107180807.GA8887@navahonetworks.com> <20050107225908.GX24460@bestpractical.com> <20050108052849.GA14618@navahonetworks.com> <20050108155716.GB16063@phlogiston.dyndns.org> <20050108210156.GE24460@bestpractical.com> Message-ID: <20050108215211.GA16581@phlogiston.dyndns.org> On Sat, Jan 08, 2005 at 04:01:56PM -0500, Jesse Vincent wrote: > > I'd love to see suggestions about how we can improve the search UI > without completely breaking it for users of downlevel browsers. (Which Well, the problem here seems more to be the number of round trips. One of my DB guys is looking at how the queries could be altered to get more in one go, but he's convinced that it needs a great deal of work to solve. > right out.) Would people be satisfied if the pick lists for users and > groups were transformed into text entry widgets? I think that'd just create a new problem, that you'd end up with groups and users that are typos. A -- Andrew Sullivan | ajs@crankycanuck.ca The plural of anecdote is not data. --Roger Brinner From jesse at fsck.com Sat Jan 8 18:55:29 2005 From: jesse at fsck.com (jesse) Date: Sat Jan 8 18:55:34 2005 Subject: [Rt-devel] RT 3.4.0rc4 Message-ID: <20050108235529.GN31352@bestpractical.com> rc4 fixes a bug in object attribute handling that could result in wonky behaviour when using the "never send mail to these people" functionality and trying to delete more than one of them at once. Also, more tests. -- From jesse at bestpractical.com Sat Jan 8 23:25:49 2005 From: jesse at bestpractical.com (Jesse Vincent) Date: Sat Jan 8 23:25:54 2005 Subject: [Rt-devel] Re: RT 3.4.0rc3 In-Reply-To: <41E09D4F.9070304@bearcircle.net> References: <20050108034321.GY24460@bestpractical.com> <41E09D4F.9070304@bearcircle.net> Message-ID: <20050109042549.GJ24460@bestpractical.com> > > Hey Jesse, > > RT 3.4.0rc3 'make testdeps' says: > > DBIx::SearchBuilder 1.19...MISSING > DBIx::SearchBuilder version 1.19 required--this is only > version 1.18 at (eval 55) line 2. > > I have rc1 installed and am trying to upgrade. I used CPAN to install > the latest version, but it only has DBIx::SearchBuilder 1.18. Is CPAN > behind? It just went up to cpan tonight. 1.18 will work reasonably well (it's only really key for the fix that rc4 fixes. > > --[Lance] > > -- > Celebrate The Circle http://www.celebratethecircle.org/ > Carolina Spirit Quest http://www.carolinaspiritquest.org/ > My LiveJournal http://www.livejournal.com/users/labrown/ > GPG Fingerprint: 409B A409 A38D 92BF 15D9 6EEE 9A82 F2AC 69AC 07B9 > CACert.org Assurer > -- From Ruslan.Zakirov at acronis.com Sun Jan 9 08:35:12 2005 From: Ruslan.Zakirov at acronis.com (Ruslan U. Zakirov) Date: Sun Jan 9 08:37:42 2005 Subject: [Rt-devel] 3.4.0rc1 slow In-Reply-To: <20050108210156.GE24460@bestpractical.com> References: <20050107180807.GA8887@navahonetworks.com> <20050107225908.GX24460@bestpractical.com> <20050108052849.GA14618@navahonetworks.com> <20050108155716.GB16063@phlogiston.dyndns.org> <20050108210156.GE24460@bestpractical.com> Message-ID: <41E13310.7080609@acronis.com> Jesse Vincent wrote: > > [snip] > > Would people be satisfied if the pick lists for users and > groups were transformed into text entry widgets? > No. Regards, Ruslan. From dkrt at navahonetworks.com Sun Jan 9 13:59:39 2005 From: dkrt at navahonetworks.com (David Kerry) Date: Sun Jan 9 14:02:04 2005 Subject: [Rt-devel] Another sql improvement recommendation for rt3.4.0 Message-ID: <20050109185939.GA29777@navahonetworks.com> Another point regarding slow database access and RT 3.4.0rc1... I've been going through the queries being run, and I think I can make one simple recommendation to improve the performance of RT significantly (having spent a fair bit of time with various flavours of databases over the years). If possible, avoid using functions on a column in a query. ie: select * from users where lower(email)='abc@bar.com'; In all the databases I've used, the optimizer will not use any indexes associated with that column as soon as you wrap it in a function and will do a full table scan. In the above example, an index defined as: "create index on users(email)" would _not_ be used for that query. There are ways around this (ie: Oracle's function-based indexes, but I don't know if that's supported by all databases and that's a bit of band-aid solution). The better way is to store the data in a known format (all lowercase in this example) and do something like this instead: select * from users where email=lower('abc@bar.com'); I point this out because loading up a ticket used this query: SELECT DISTINCT main.* FROM Tickets main, Groups Groups_1, CachedGroupMembers CachedGroupMembers_2, Users Users_3 WHERE (CachedGroupMembers_2.MemberId = Users_3.id) AND (Groups_1.id = CachedGroupMembers_2.GroupId) AND (main.EffectiveId = main.id) AND (main.Status != 'deleted') AND (main.id = Groups_1.Instance) AND (((LOWER(Users_3.EmailAddress) = 'user@abc.com') AND (Groups_1.Type = 'Requestor')) AND ((main.Status ='new') OR (main.Status ='open'))) ORDER BY main.Priority DESC LIMIT 10; It took close to 2 seconds to run on my database (27k users). Remove the 'LOWER(Users_3.EmailAddress)' and the query takes < 10ms because it uses all the indexes. -- David Kerry From jesse at bestpractical.com Sun Jan 9 14:50:16 2005 From: jesse at bestpractical.com (Jesse Vincent) Date: Sun Jan 9 14:50:21 2005 Subject: [Rt-devel] Another sql improvement recommendation for rt3.4.0 In-Reply-To: <20050109185939.GA29777@navahonetworks.com> References: <20050109185939.GA29777@navahonetworks.com> Message-ID: <20050109195015.GO24460@bestpractical.com> > > I've been going through the queries being run, and I think I can > make one simple recommendation to improve the performance of > RT significantly (having spent a fair bit of time with various > flavours of databases over the years). > > If possible, avoid using functions on a column in a query. If you look a bit deeper, you'll see that that's being done by SearchBuilder when a case-insensitive search is called for on a database that doesn't natively support case-insensitive queries. > In all the databases I've used, the optimizer will not use > any indexes associated with that column as soon as you wrap > it in a function and will do a full table scan. Postgres and Oracle both support functional indexes. > The better way is to store the data in a known format (all lowercase > in this example) and do something like this instead: Except that would mean storing usernames in all lowercase, as well as all correspondence related to a ticket. > > > I point this out because loading up a ticket used this query: That query sure looks like a search for all tickets with user@abc.com, not loading a single ticket...What happens if you add a functional index? > SELECT DISTINCT main.* > FROM Tickets main, Groups Groups_1, CachedGroupMembers CachedGroupMembers_2, > Users Users_3 > WHERE (CachedGroupMembers_2.MemberId = Users_3.id) AND > (Groups_1.id = CachedGroupMembers_2.GroupId) AND > (main.EffectiveId = main.id) AND > (main.Status != 'deleted') AND > (main.id = Groups_1.Instance) AND > (((LOWER(Users_3.EmailAddress) = 'user@abc.com') AND > (Groups_1.Type = 'Requestor')) AND > ((main.Status ='new') OR (main.Status ='open'))) > ORDER BY main.Priority DESC LIMIT 10; > > It took close to 2 seconds to run on my database (27k users). > > Remove the 'LOWER(Users_3.EmailAddress)' and the query takes < 10ms because > it uses all the indexes. > > -- > David Kerry > > _______________________________________________ > Rt-devel mailing list > Rt-devel@lists.bestpractical.com > http://lists.bestpractical.com/cgi-bin/mailman/listinfo/rt-devel > -- From dkrt at navahonetworks.com Sun Jan 9 15:31:50 2005 From: dkrt at navahonetworks.com (David Kerry) Date: Sun Jan 9 15:34:17 2005 Subject: [Rt-devel] Another sql improvement recommendation for rt3.4.0 In-Reply-To: <20050109195015.GO24460@bestpractical.com> References: <20050109185939.GA29777@navahonetworks.com> <20050109195015.GO24460@bestpractical.com> Message-ID: <20050109203150.GA30242@navahonetworks.com> On Sun, Jan 09, 2005 at 02:50:16PM -0500, Jesse Vincent wrote: > > > If possible, avoid using functions on a column in a query. > > If you look a bit deeper, you'll see that that's being done by > SearchBuilder when a case-insensitive search is called for on a database > that doesn't natively support case-insensitive queries. Ah.. ok. > > In all the databases I've used, the optimizer will not use > > any indexes associated with that column as soon as you wrap > > it in a function and will do a full table scan. > > Postgres and Oracle both support functional indexes. True - but it's nice to avoid such things if they're not really necessary. > > The better way is to store the data in a known format (all lowercase > > in this example) and do something like this instead: > > Except that would mean storing usernames in all lowercase, as well as > all correspondence related to a ticket. I didn't mean all columns had to be done this way, just cases where it can be done easily. For example - I see no reason to not store user's email addresses in all-lowercase. Email addresses are naturally case-insensitive anyways, so no problems there. That would remove the one slow query I found already. > > I point this out because loading up a ticket used this query: > > That query sure looks like a search for all tickets with user@abc.com, > not loading a single ticket...What happens if you add a functional > index? This was for loading up a single ticket - I presume it was looking for any other tickets created by the same user to list in one of the related ticket display boxes. For what it's worth, adding a function based index didn't seem to make any difference in postgres. ie: create index users5dk on users(lower(emailaddress)); Query time was the same (2-4 seconds) with or without this index. I've also found another one: SELECT * FROM Groups WHERE LOWER(Domain) = 'systeminternal' AND LOWER(Type) = 'privileged'; 500ms (with LOWER()) vs 36 ms (without LOWER()). This seems to be an internal query - RT should know the values in its own internal tables, no? Why the case-insensitive search for this? If the lower() is being added to columns automatically, I think this is probably a major cause of some of the RT slowness. Overzealous use of that is going to kill your performance (unless you make all your indexes case-insensitive function-based indexes, that is). Also, note that Oracle (haven't tried postgres) won't allow you to index a column more than once (with a single-column index), so you either have a case-sensitive index (lower()), or a case-insensitive one, but you can't have both, so you either have to do all your queries with a matching lower() or none at all (or as few as possible). Anyways... I'm just pointing out that you can get some major performance boosts out of RT if you fine tune some of the queries a bit. -- David Kerry From dschmidt at noanet.net Sun Jan 9 15:34:28 2005 From: dschmidt at noanet.net (David Schmidt) Date: Sun Jan 9 15:36:51 2005 Subject: [Rt-devel] Problems converting rt-2.0.15 to rt-3.0.9 using rt-2-to-rt3 Message-ID: <6B628149FCE1B543A3AB351A29DC7D8E471350@basilisk.noanet.lan> I am attempting to convert a database from rt-2-0-15 to rt-3-0-9 using rt-2-to-rt3-1.23 and getting multiple errors. We are running mysql-max-4.1.7 under FreeBSD 4.10. I'm new to RT and haven't worked with Perl in many years, so feel free to correct any errors below. The rt-2.0-to-dumpfile runs fine, but I'm getting errors when I run the dumpfile-to-rt-3.0 script. I am doing a "make dropdb" and "make initialize-database" on the RT3 database before every run of dumpfile-to-rt-3.0. The first error was in importing users, and seems to have been caused by "undef" values being sent to SearchBuilder. Around line 222 I found the following code: if ($args{'CryptedPassword'} ) { $args{'Password'} = $args{'CryptedPassword'}; delete $args{'CryptedPassword'}; } elsif ( !$args{'Password'} ) { $args{'Password'} = '*NO-PASSWORD*'; } elsif ( length( $args{'Password'} ) < $RT::MinimumPasswordLength ) { return ( 0, $self->loc("Password too short") ); } else { $args{'Password'} = $self->_GeneratePassword($args{'Password'}); } If CryptedPassword is undefined the first if will fail, setting Password to "*NO-PASSWORD*", but NOT deleting the key for CryptedPassword. I changed this code to delete the CryptedPassword no matter which branch is taken, and then to perform the following tests for password length. if ($args{'CryptedPassword'} ) { $args{'Password'} = $args{'CryptedPassword'}; } elsif ( !$args{'Password'} ) { $args{'Password'} = '*NO-PASSWORD*'; } delete $args{'CryptedPassword'}; This got the user names imported, but now I'm getting the following errors after creating all the users: Importing groups gggrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrr rrrrrfqF ailed to create queue for$VAR1 = { 'LastUpdatedBy' => undef, 'Creator' => undef }; Couldn't load principal to grant them AdminKeywordSelects on queue $VAR1 = { 'PrincipalType' => 'Group', 'Role' => 'AdminCc', 'Name' => 'AdminKeywordSelects' }; Couldn't load principal to grant them AdminQueue on queue $VAR1 = { 'PrincipalType' => 'Group', 'Role' => 'AdminCc', 'Name' => 'AdminQueue' }; Couldn't load principal to grant them CommentOnTicket on queue I was wondering if anyone else has had similar problems and come up with a fix or updated rt-2-to-rt3? Thank you, David Schmidt NOC Analyst Northwest Open Access Network 1-866-662-6380 dschmidt@noanet.net -------------- next part -------------- An HTML attachment was scrubbed... URL: http://bestpractical.com/pipermail/rt-devel/attachments/20050109/8c93acf5/attachment-0001.htm From jesse at bestpractical.com Sun Jan 9 15:43:15 2005 From: jesse at bestpractical.com (Jesse Vincent) Date: Sun Jan 9 15:43:20 2005 Subject: [Rt-devel] Another sql improvement recommendation for rt3.4.0 In-Reply-To: <20050109203150.GA30242@navahonetworks.com> References: <20050109185939.GA29777@navahonetworks.com> <20050109195015.GO24460@bestpractical.com> <20050109203150.GA30242@navahonetworks.com> Message-ID: <20050109204315.GQ24460@bestpractical.com> > > Except that would mean storing usernames in all lowercase, as well as > > all correspondence related to a ticket. > > I didn't mean all columns had to be done this way, just cases where > it can be done easily. For example - I see no reason to not store > user's email addresses in all-lowercase. Email addresses are naturally > case-insensitive anyways, so no problems there. That would remove > the one slow query I found already. No, they're not. You may not use a mail system that's case sensitive. But email address local parts are, at the option of the local system, case sensitive. From RFC 2821: Mailbox = Local-part "@" Domain Local-part = Dot-string / Quoted-string ; MAY be case-sensitive > I've also found another one: > > SELECT * FROM Groups WHERE LOWER(Domain) = 'systeminternal' AND LOWER(Type) = 'privileged'; > > 500ms (with LOWER()) vs 36 ms (without LOWER()). > > This seems to be an internal query - RT should know the values in its > own internal tables, no? Why the case-insensitive search for this? Likely it needs to be hinted. Open a ticket, ideally with a patch that adds CASESENITIVE => 1, to the calls to seachbuilder->Limit. -- From dkrt at navahonetworks.com Sun Jan 9 15:43:17 2005 From: dkrt at navahonetworks.com (David Kerry) Date: Sun Jan 9 15:45:44 2005 Subject: [Rt-devel] 3.4.0rc1 slow In-Reply-To: <20050108062157.GC24460@bestpractical.com> References: <20050107180807.GA8887@navahonetworks.com> <20050107225908.GX24460@bestpractical.com> <20050108052849.GA14618@navahonetworks.com> <20050108062157.GC24460@bestpractical.com> Message-ID: <20050109204317.GB30242@navahonetworks.com> On Sat, Jan 08, 2005 at 01:21:57AM -0500, Jesse Vincent wrote: > > > > > Simply bringing up the tickets->query builder page generates over 129 sql > > queries ?! 19 of those queries took between 400 and 600ms apiece. That's > > an awful lot of data trolling to bring up one screen. Is that right, or > > is my setup misbehaving/borked somehow? I have saved the entire sql trace > > in case anyone is interested in looking at it in detail. > > > > My first guess is that it's mostly in the dropdowns for picking ticket > owners by name or group membership. Using MasonX::Profiler, you might > be able to see which components are causing the slowdowns, which will > help to find out how to fix em. Was 3.2 similarly slow? I haven't used 3.2 yet. I'm considering making the big leap from 2.0.13 direct to 3.4 (due to growing complaints from my users over the slowness of 2.0). I'll play with MasonX::Profile and see what I can find out as well. -- David Kerry From Brian_Gupta at timeinc.com Sun Jan 9 15:28:20 2005 From: Brian_Gupta at timeinc.com (Brian_Gupta@timeinc.com) Date: Sun Jan 9 15:54:24 2005 Subject: [Rt-devel] RE: Rt-devel Digest, Vol 10, Issue 7 Message-ID: <4D0FD2F4B17F334AAC95B9CE02800477362B2F@nycmailmb04.enterprise.corpad.timeinc.com> First let me say, I can't program, nor do I fully understand the RT architecture... That said, I was wondering if any of the following ideas make sense: Could the query builder page and pick lists, be preloaded via the mason interpreter under modperl? (There seems to a mason option) One could also try using mason caching for the the picklists, (Mason does have a caching mechanism) We would just have to add cache control code to the RT GUI areas that could modify the picklist entries... Brian Gupta Time Inc Information Technology Dept 212-522-1401 Message: 2 Date: Sat, 8 Jan 2005 16:52:11 -0500 From: Andrew Sullivan Subject: Re: [Rt-devel] 3.4.0rc1 slow To: rt-devel@lists.bestpractical.com Message-ID: <20050108215211.GA16581@phlogiston.dyndns.org> Content-Type: text/plain; charset=us-ascii On Sat, Jan 08, 2005 at 04:01:56PM -0500, Jesse Vincent wrote: > > I'd love to see suggestions about how we can improve the search UI > without completely breaking it for users of downlevel browsers. (Which Well, the problem here seems more to be the number of round trips. One of my DB guys is looking at how the queries could be altered to get more in one go, but he's convinced that it needs a great deal of work to solve. > right out.) Would people be satisfied if the pick lists for users and > groups were transformed into text entry widgets? I think that'd just create a new problem, that you'd end up with groups and users that are typos. A -- Andrew Sullivan | ajs@crankycanuck.ca The plural of anecdote is not data. --Roger Brinner ------------------------------ From jesse at bestpractical.com Sun Jan 9 16:14:56 2005 From: jesse at bestpractical.com (Jesse Vincent) Date: Sun Jan 9 16:15:01 2005 Subject: [Rt-devel] RE: Rt-devel Digest, Vol 10, Issue 7 In-Reply-To: <4D0FD2F4B17F334AAC95B9CE02800477362B2F@nycmailmb04.enterprise.corpad.timeinc.com> References: <4D0FD2F4B17F334AAC95B9CE02800477362B2F@nycmailmb04.enterprise.corpad.timeinc.com> Message-ID: <20050109211455.GS24460@bestpractical.com> On Sun, Jan 09, 2005 at 03:28:20PM -0500, Brian_Gupta@timeinc.com wrote: > First let me say, I can't program, nor do I fully understand the RT > architecture... That said, I was wondering if any of the following ideas > make sense: > > Could the query builder page and pick lists, be preloaded via the mason > interpreter under modperl? (There seems to a mason option) > > One could also try using mason caching for the the picklists, (Mason > does have a caching mechanism) We would just have to add cache control > code to the RT GUI areas that could modify the picklist entries... Part of the problem is that those lists vary (or should be varying) based on the queues you have selected and the permissions you have right now. They're supposed to be smart and dynamic and that hurts, performancewise. From Brian_Gupta at timeinc.com Sun Jan 9 18:18:49 2005 From: Brian_Gupta at timeinc.com (Brian_Gupta@timeinc.com) Date: Sun Jan 9 18:19:49 2005 Subject: [Rt-devel] RE: Rt-devel Digest, Vol 10, Issue 7 Message-ID: <4D0FD2F4B17F334AAC95B9CE028004773C6E51@nycmailmb04.enterprise.corpad.timeinc.com> Does RT currently take advantage of Apache sessions via cookies? The MASON caching directives can be done on a per session basis. (If you do something like log out or change the configurations, the code would need to be added to should clear the session cache(s)) Brian Gupta Time Inc Information Technology Dept 212-522-1401 -----Original Message----- From: Jesse Vincent [mailto:jesse@bestpractical.com] Sent: Sunday, January 09, 2005 4:15 PM To: Gupta, Brian - Information Technology Cc: rt-devel@lists.bestpractical.com Subject: Re: [Rt-devel] RE: Rt-devel Digest, Vol 10, Issue 7 On Sun, Jan 09, 2005 at 03:28:20PM -0500, Brian_Gupta@timeinc.com wrote: > First let me say, I can't program, nor do I fully understand the RT > architecture... That said, I was wondering if any of the following > ideas make sense: > > Could the query builder page and pick lists, be preloaded via the > mason interpreter under modperl? (There seems to a mason option) > > One could also try using mason caching for the the picklists, (Mason > does have a caching mechanism) We would just have to add cache control > code to the RT GUI areas that could modify the picklist entries... Part of the problem is that those lists vary (or should be varying) based on the queues you have selected and the permissions you have right now. They're supposed to be smart and dynamic and that hurts, performancewise. From matthew at trebex.net Sun Jan 9 20:04:13 2005 From: matthew at trebex.net (Matthew Draper) Date: Sun Jan 9 20:13:14 2005 Subject: [Rt-devel] RE: Rt-devel Digest, Vol 10, Issue 7 In-Reply-To: <4D0FD2F4B17F334AAC95B9CE028004773C6E51@nycmailmb04.enterprise.corpad.timeinc.com> References: <4D0FD2F4B17F334AAC95B9CE028004773C6E51@nycmailmb04.enterprise.corpad.timeinc.com> Message-ID: <41E1D48D.9070207@trebex.net> Brian_Gupta@timeinc.com wrote: > The MASON caching directives can be done on a per session basis. That doesn't help much; the list content changes not just on the current user's permissions, but also on the context of the list... when setting an owner, the list only shows users that are permitted to own the particular ticket being modified, for example. The memory required to cache a list of potential owners for every ticket viewed in a session makes such caching somewhat less than entirely practical. -- matthew@trebex.net -------------- next part -------------- A non-text attachment was scrubbed... Name: signature.asc Type: application/pgp-signature Size: 254 bytes Desc: OpenPGP digital signature Url : http://bestpractical.com/pipermail/rt-devel/attachments/20050110/bc41f15d/signature.pgp From vivek at khera.org Sun Jan 9 21:26:48 2005 From: vivek at khera.org (Vivek Khera) Date: Sun Jan 9 21:53:50 2005 Subject: [Rt-devel] 3.4.0rc1 slow In-Reply-To: <20050107225908.GX24460@bestpractical.com> References: <20050107180807.GA8887@navahonetworks.com> <20050107225908.GX24460@bestpractical.com> Message-ID: <14BB9484-62AF-11D9-B522-000A95D14982@khera.org> On Jan 7, 2005, at 5:59 PM, Jesse Vincent wrote: >> postgres[18131]: [6610-1] LOG: duration: 881.154 ms statement: >> SELECT id FROM Attachments WHERE oid = '3609306' >> >> From doing a bit of reading in the postgres docs, the oid should >> not >> be used for this type of thing - they can/will roll, which can lead >> to duplicates too. > > But not instantly. And the only time we're doing that is immediately > after the insert. > also, there's no index on the oid's so it could result in a long table scan for a large attachments table. -------------- next part -------------- A non-text attachment was scrubbed... Name: smime.p7s Type: application/pkcs7-signature Size: 2476 bytes Desc: not available Url : http://bestpractical.com/pipermail/rt-devel/attachments/20050109/1b4fcd0e/smime.bin From ajs at crankycanuck.ca Mon Jan 10 00:26:55 2005 From: ajs at crankycanuck.ca (Andrew Sullivan) Date: Mon Jan 10 00:29:45 2005 Subject: [Rt-devel] Another sql improvement recommendation for rt3.4.0 In-Reply-To: <20050109203150.GA30242@navahonetworks.com> References: <20050109185939.GA29777@navahonetworks.com> <20050109195015.GO24460@bestpractical.com> <20050109203150.GA30242@navahonetworks.com> Message-ID: <20050110052655.GB18085@phlogiston.dyndns.org> On Sun, Jan 09, 2005 at 03:31:50PM -0500, David Kerry wrote: > > Postgres and Oracle both support functional indexes. > > True - but it's nice to avoid such things if they're not really > necessary. Why? First, they _are_ necessary in this case (although I confess that I have the distinct impression some cases are because of MySQL's non-standard understanding of "=" in SQL). But second, why would anyone implement functional indexes when they're not really necessary? A -- Andrew Sullivan | ajs@crankycanuck.ca A certain description of men are for getting out of debt, yet are against all taxes for raising money to pay it off. --Alexander Hamilton From mhat at netlag.com Mon Jan 10 00:53:35 2005 From: mhat at netlag.com (Matt Knopp) Date: Mon Jan 10 01:20:07 2005 Subject: [Rt-devel] Another sql improvement recommendation for rt3.4.0 In-Reply-To: <20050110052655.GB18085@phlogiston.dyndns.org> References: <20050109185939.GA29777@navahonetworks.com> <20050109195015.GO24460@bestpractical.com> <20050109203150.GA30242@navahonetworks.com> <20050110052655.GB18085@phlogiston.dyndns.org> Message-ID: <20050110055335.GD942@netlag.com> Umm, are you seriously asking why a software company would add a feature to their software that may or may not really be needed?? *Blink* *Blink* -- Matt > But second, why would anyone implement functional indexes when > they're not really necessary? From ajs at crankycanuck.ca Mon Jan 10 03:49:18 2005 From: ajs at crankycanuck.ca (Andrew Sullivan) Date: Mon Jan 10 03:52:10 2005 Subject: [Rt-devel] Another sql improvement recommendation for rt3.4.0 In-Reply-To: <20050110055335.GD942@netlag.com> References: <20050109185939.GA29777@navahonetworks.com> <20050109195015.GO24460@bestpractical.com> <20050109203150.GA30242@navahonetworks.com> <20050110052655.GB18085@phlogiston.dyndns.org> <20050110055335.GD942@netlag.com> Message-ID: <20050110084918.GC18085@phlogiston.dyndns.org> On Sun, Jan 09, 2005 at 11:53:35PM -0600, Matt Knopp wrote: > > Umm, are you seriously asking why a software company would add a > feature to their software that may or may not really be needed?? > Given that PostgreSQL is not "a software company", yes, I am asking that. A -- Andrew Sullivan | ajs@crankycanuck.ca Information security isn't a technological problem. It's an economics problem. --Bruce Schneier From rt at tmtm.com Mon Jan 10 14:10:44 2005 From: rt at tmtm.com (Tony Bowden) Date: Mon Jan 10 14:44:07 2005 Subject: [Rt-devel] 3.4.0rc1 slow In-Reply-To: <20050108210156.GE24460@bestpractical.com> References: <20050107180807.GA8887@navahonetworks.com> <20050107225908.GX24460@bestpractical.com> <20050108052849.GA14618@navahonetworks.com> <20050108155716.GB16063@phlogiston.dyndns.org> <20050108210156.GE24460@bestpractical.com> Message-ID: <20050110191044.GA5429@soto.kasei.com> On Sat, Jan 08, 2005 at 04:01:56PM -0500, Jesse Vincent wrote: > I'd love to see suggestions about how we can improve the search UI > without completely breaking it for users of downlevel browsers. (Which > means that pretty javascript hacks and popup windows are pretty much > right out.) Why does it rule out Javascript. What's wrong with having something fancier for people who have JS with the fall back to the clumsier approach for people who don't. Surely JS should only be avoided if it's the *only* way to achieve something? Tony From dschmidt at noanet.net Mon Jan 10 15:30:03 2005 From: dschmidt at noanet.net (David Schmidt) Date: Mon Jan 10 15:32:38 2005 Subject: [Rt-devel] Problems converting rt-2.0.15 to rt-3.0.9 using rt-2-to-rt3 Message-ID: <6B628149FCE1B543A3AB351A29DC7D8E471439@basilisk.noanet.lan> I have done some additional debugging on the error I'm currently getting: This got the user names imported, but now I'm getting the following errors after creating all the users: Importing groups gggrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrr rrrrrfqF ailed to create queue for$VAR1 = { 'LastUpdatedBy' => undef, 'Creator' => undef }; Using the Perl debugger, I traced the "$queue_obj->Create( %{$queue} );" call on line 246 of dumpfile-to-rt-3.0 and found that the problem was that the db INSERT failed because there was no value supplied for the "Name" column. Where should that Name value be coming from? David Schmidt NOC Analyst Northwest Open Access Network dschmidt@noanet.net 1-866-662-6380 From mrz at intelenet.net Mon Jan 10 19:35:45 2005 From: mrz at intelenet.net (matthew zeier) Date: Mon Jan 10 19:59:37 2005 Subject: [Rt-devel] integrating RT with third-party apps ? Message-ID: <20050111003545.GB11528@intelenet.net> Is there a 'simple' method to integrate RT into other apps? More in the line of some other app interacting with RT to extract data and make ticket changes (XML)? -- matthew zeier | "In mathematics you don't understand InteleNet Communications, Inc. | things. You just get used to them." (949) 784-7904 | - John von Newmann From jralbert at uvic.ca Mon Jan 10 23:12:30 2005 From: jralbert at uvic.ca (Jeff Albert) Date: Tue Jan 11 00:15:18 2005 Subject: [Rt-devel] Ticket Ownership Enumeration Problem Message-ID: <0IA400E4SX0TJV@l-daemon> Hello all, I'm a beginner at hacking RT, and at Perl, so please forgive me if this turns out to be a foolish question, but: I'm trying to create a scrip which will, upon ticket creation, evaluate a custom field value to determine which group of our team should address the issue, then determine who among them has the fewest open tickets, and finally choose a random member of that subset to be assigned ownership of the ticket. Presently I'm stuck at trying to determine how many tickets a given user has open. It seems to me that this should be fairly trivial, but I'm having no luck. Here's what I'm doing currently: #...$org is gotten from custom field... if ($org eq 'HUMA') { my %HSD = ('jralbert' => 0, 'jafraser'=> 0, 'mhelfric' => 0); my $CurrentUser = $RT::SystemUser; my $OpenTickets; $OpenTickets = new RT::Tickets ($CurrentUser); while ($id, $score) = each(%HSD)) { $MyTickets->UnLimit(); $MyTickets->LimitOwner( VALUE => "jralbert" ); $MyTickets->LimitStatus( VALUE=> "open" ); $HSD{$id} = $MyTickets->Count(); } } It looks as if the scrip dies on trying the LimitOwner method, and gets no further. What am I doing wrong? I thank you in advance for any help you may provide. Cheers, Jeff Albert Departmental Computing Support Team University of Victoria British Columbia, Canada jralbert at uvic dot ca From jesse at bestpractical.com Tue Jan 11 00:28:29 2005 From: jesse at bestpractical.com (Jesse Vincent) Date: Tue Jan 11 00:28:34 2005 Subject: [Rt-devel] Ticket Ownership Enumeration Problem In-Reply-To: <0IA400E4SX0TJV@l-daemon> References: <0IA400E4SX0TJV@l-daemon> Message-ID: <20050111052829.GG24460@bestpractical.com> > #...$org is gotten from custom field... > > if ($org eq 'HUMA') { > my %HSD = ('jralbert' => 0, 'jafraser'=> 0, 'mhelfric' => 0); > my $CurrentUser = $RT::SystemUser; > my $OpenTickets; > $OpenTickets = new RT::Tickets ($CurrentUser); # This line serves no purpose > while ($id, $score) = each(%HSD)) { # There's no MyTickets object > $MyTickets->UnLimit(); # Rather than do this, why don't you do: my $MyTickets = RT::Tickets->new($RT::SystemUser); > $MyTickets->LimitOwner( VALUE => "jralbert" ); > $MyTickets->LimitStatus( VALUE=> "open" ); > $HSD{$id} = $MyTickets->Count(); > } > } > Once you have this all worked up, please document it on the wiki. Jesse > It looks as if the scrip dies on trying the LimitOwner method, and gets no > further. What am I doing wrong? I thank you in advance for any help you may > provide. > > Cheers, > Jeff Albert > > Departmental Computing Support Team > University of Victoria > British Columbia, Canada > jralbert at uvic dot ca > > > _______________________________________________ > Rt-devel mailing list > Rt-devel@lists.bestpractical.com > http://lists.bestpractical.com/cgi-bin/mailman/listinfo/rt-devel > -- From jralbert at uvic.ca Tue Jan 11 01:53:01 2005 From: jralbert at uvic.ca (Jeff Albert) Date: Tue Jan 11 01:55:59 2005 Subject: [Rt-devel] Ticket Ownership Enumeration Problem In-Reply-To: <20050111052829.GG24460@bestpractical.com> Message-ID: <0IA50017N4GA24@l-daemon> Yes, the code was an error-riddled mess. Sorry about that! I really shouldn't post while halfway through altering my code. Here's the revised, and now-working version: #org is gotten from custom field... if ($org eq 'HUMA') { #define team and vars... my %HSD = ('jralbert' => 0, 'jafraser'=> 0, 'mhelfric' => 0); my $CurrentUser = $RT::SystemUser; my $id; my $score; #now, let's find out how many tickets everyone has. while (($id, $score) = each(%HSD)) { my $OpenTickets = new RT::Tickets ($CurrentUser); $OpenTickets->LimitOwner( VALUE => $id ); $OpenTickets->LimitQueue( VALUE => 'COUS_DCST_Test' ); $OpenTickets->LimitStatus ( VALUE => 'open' ); $HSD{$id} = $OpenTickets->Count(); $report .= ' ' . $id . ' : ' . $HSD{$id}; } #next, let's set a threshold value absurdly high, #then whittle it down to the lowest number of tickets any user has open. my $topscore = 999; while (($id, $score) = each(%HSD)) { if ($score < $topscore) { $topscore = $score; } } #finally, let's get subset of our team who have that number of tickets open... my @candidates; while (($id, $score) = each(%HSD)) { if ($score == $topscore){ push(@candidates, $id); } } #...and pick one of them randomly to be assigned the task. my $final = $candidates[int rand scalar @candidates]; $self->TicketObj->SetOwner($final); } If this code is useful to anyone, they're welcome to it! Thanks for the feedback and help! Cheers, Jeff Albert Departmental Computing Support Team University of Victoria British Columbia, Canada jralbert at uvic dot ca -----Original Message----- From: Jesse Vincent [mailto:jesse@bestpractical.com] Sent: Monday, January 10, 2005 9:28 PM To: Jeff Albert Cc: rt-devel@lists.bestpractical.com Subject: Re: [Rt-devel] Ticket Ownership Enumeration Problem > #...$org is gotten from custom field... > > if ($org eq 'HUMA') { > my %HSD = ('jralbert' => 0, 'jafraser'=> 0, 'mhelfric' => 0); > my $CurrentUser = $RT::SystemUser; > my $OpenTickets; > $OpenTickets = new RT::Tickets ($CurrentUser); # This line serves no purpose > while ($id, $score) = each(%HSD)) { # There's no MyTickets object > $MyTickets->UnLimit(); # Rather than do this, why don't you do: my $MyTickets = RT::Tickets->new($RT::SystemUser); > $MyTickets->LimitOwner( VALUE => "jralbert" ); > $MyTickets->LimitStatus( VALUE=> "open" ); > $HSD{$id} = $MyTickets->Count(); > } > } > Once you have this all worked up, please document it on the wiki. Jesse > It looks as if the scrip dies on trying the LimitOwner method, and gets no > further. What am I doing wrong? I thank you in advance for any help you may > provide. > > Cheers, > Jeff Albert > > Departmental Computing Support Team > University of Victoria > British Columbia, Canada > jralbert at uvic dot ca > > > _______________________________________________ > Rt-devel mailing list > Rt-devel@lists.bestpractical.com > http://lists.bestpractical.com/cgi-bin/mailman/listinfo/rt-devel > -- From nordstrom at realnode.com Tue Jan 11 04:38:56 2005 From: nordstrom at realnode.com (Mattias Nordstrom) Date: Tue Jan 11 04:41:32 2005 Subject: [Rt-devel] Custom Status values Message-ID: <41E39EB0.6080300@realnode.com> Hi, Is there an easy way to add new Status values? Or are they hard coded in many files? TIA -- Mattias Nordstrom CTO Oy Realnode Ab nordstrom@realnode.com www.realnode.com From Ruslan.Zakirov at acronis.com Tue Jan 11 08:04:00 2005 From: Ruslan.Zakirov at acronis.com (Ruslan U. Zakirov) Date: Tue Jan 11 08:26:26 2005 Subject: [Rt-devel] integrating RT with third-party apps ? In-Reply-To: <20050111003545.GB11528@intelenet.net> References: <20050111003545.GB11528@intelenet.net> Message-ID: <41E3CEC0.1040901@acronis.com> 1) Atom interface and here is client for it: http://search.cpan.org/~autrijus/RT-Client-0.01/ 2) REST API 3) RT API 4) mailgate to create new tickets matthew zeier wrote: > Is there a 'simple' method to integrate RT into other apps? More in the > line of some other app interacting with RT to extract data and make > ticket changes (XML)? > From Ruslan.Zakirov at acronis.com Tue Jan 11 08:42:43 2005 From: Ruslan.Zakirov at acronis.com (Ruslan U. Zakirov) Date: Tue Jan 11 08:45:29 2005 Subject: [Rt-devel] Custom Status values In-Reply-To: <41E39EB0.6080300@realnode.com> References: <41E39EB0.6080300@realnode.com> Message-ID: <41E3D7D3.4070807@acronis.com> Use subject of your message as search condition in ML archives http://wiki.bestpractical.com/?MailingListArchives Mattias Nordstrom wrote: > Hi, > > Is there an easy way to add new Status values? Or are they hard coded in > many files? > > TIA > From joel48 at ischool.washington.edu Tue Jan 11 18:44:23 2005 From: joel48 at ischool.washington.edu (Joel Johnson) Date: Tue Jan 11 19:25:11 2005 Subject: [Rt-devel] RT-RTFM Integration Message-ID: <41E464D7.3050707@ischool.washington.edu> We're using RT 3.2.2 (debian package) and RTFM 2.0.4 (tarball install, kudos to Stephen to the link to the RTFM .debs a few days ago). Working great, except for a few integration issues between the two. When modifying a ticket and I do a search for RTFM articles by content, RTFM/Ticket/Update.html/BeforeMessageBox runs two while loops that iterate sequentially through $articles_content, followed by searching $articles_basics, merrily dumping HTML as they go. As a result, if a search term matches the Name and content of a FAQ, duplicates are returned (with the same "RTFM-Include-Article-<%ID%>" for each). Being relatively (9 months) new to RTFM, I'm wondering if there is a reason for this behaviour, or if it would be better to cache the searches into a data structure, then remove duplicates, and dump the HTML lines from the data structure. How would this scale with a large number of FAQs? For that matter, what is the largest number of FAQs in one installation people are aware of? FAQs would seem to be to be a more static set then tickets, which are always growing. The other issue that has always bothered me is the setup of the Tabs between RT and RTFM. In RT, RTFM shows up as any other tab, so why does RTFM relegate RT to a single tab, isolating itself? I think it would offer a more unified approach to the users if the RTFM tab opened a submenu, ala Tickets, maintaining the rest of the navigational structure. I have only quickly glanced at the code on this one, however... has this idea already been hashed over? Rationale for the current organization? On another note, how much active RTFM development is taking place? Is there a subversion repository? Thanks, Joel From jesse at bestpractical.com Tue Jan 11 20:53:08 2005 From: jesse at bestpractical.com (Jesse Vincent) Date: Tue Jan 11 20:53:14 2005 Subject: [Rt-devel] RT-RTFM Integration In-Reply-To: <41E464D7.3050707@ischool.washington.edu> References: <41E464D7.3050707@ischool.washington.edu> Message-ID: <20050112015308.GM24460@bestpractical.com> > returned (with the same "RTFM-Include-Article-<%ID%>" for each). Being > relatively (9 months) new to RTFM, I'm wondering if there is a reason > for this behaviour, or if it would be better to cache the searches into > a data structure, then remove duplicates, and dump the HTML lines from > the data structure. How would this scale with a large number of FAQs? Nope. Patches would be appreciated. > The other issue that has always bothered me is the setup of the Tabs > between RT and RTFM. In RT, RTFM shows up as any other tab, so why does > RTFM relegate RT to a single tab, isolating itself? I think it would > offer a more unified approach to the users if the RTFM tab opened a > submenu, ala Tickets, maintaining the rest of the navigational > structure. I have only quickly glanced at the code on this one, > however... has this idea already been hashed over? Rationale for the > current organization? No specific reason. I'd love to see an experimental patch for this to see how it would work. > On another note, how much active RTFM development is taking place? Is > there a subversion repository? It's in the same svn repository as RT. Things have been a bit quiet of late. There's a 2.1 development branch that wants to become 2.2 soon. We've also got an upcoming customer project to do some more interesting new RTFM work. Jesse > Thanks, > Joel > _______________________________________________ > Rt-devel mailing list > Rt-devel@lists.bestpractical.com > http://lists.bestpractical.com/cgi-bin/mailman/listinfo/rt-devel > -- From prouss at cae.com Thu Jan 13 13:49:27 2005 From: prouss at cae.com (Patrick Rousseau-Mathieu) Date: Thu Jan 13 13:52:38 2005 Subject: [Rt-devel] Content type, in the history Message-ID: I want to know if someone know where's the content type of the email receive via rt-mailgate is set (in the code) when we want to watch the message in the history of the create ticket. I'd like to put this to: text/html and not text/plain. Tnx From paul at marchex.com Thu Jan 13 14:08:07 2005 From: paul at marchex.com (Paul Petersen) Date: Thu Jan 13 14:17:44 2005 Subject: [Rt-devel] Requestor OR email like query from SearchBuilder sticks in mysql- RT v3.2.2 Message-ID: <1105643287.3037.128.camel@blades.marchex.com> There is a query built through the SearchBuilder that consistently hangs the query in mysql. I searched the mail list archives and the wiki and only found similar situations that dealt with watchers and was fixed in v3+ I'm hoping someone can reproduce to verify it is not an issue with my setup and I apologize if I've missed prior reports of this. The particulars- RT v 3.2.2 Mysql v 4.0.15 Apache v 2.0.45 Perl 5.8.5 I've attached the config output from the RT System Configuration for any other needed details. To reproduce- Go to Tickets, New Query- Select OR radio button Select Owner is (some user in system) Select Requestor EmailAddress contains (part of email address of above person) Variations of the above for the Requestor (RealName, is, etc.) produce the same results. The query that gets stuck on my system looks like this in the process list: | 355513 | rt_user | localhost | rt3 | Query | 63338 | Sending data | SELECT COUNT(DISTINCT main.id) FROM Tickets main , Groups Groups_1, CachedGroupMembers CachedGroupMembers_2, Users Users_3 WHERE ((main.EffectiveId = main.id)) AND ((main.Status != 'deleted')) AND ( ( (Users_3.EmailAddress LIKE '%ajcantu%')AND(Groups_1.Domain = 'RT::Ticket-Role')AND(main.id = Groups_1.Instance)AND(Groups_1.Type = 'Requestor')AND(Groups_1.id = CachedGroupMembers_2.GroupId)AND(CachedGroupMembers_2.MemberId = Users_3.id) ) OR(main.Owner = '414')) | The 63338 is the time it has been running and they never finish. I found this after a user kept trying to do a search finding tickets he owned and requested. There were so many of these that mysql finally became unrepsonsive to RT and RT started to 500 waiting for it. I've had to manually kill these processes to get rid of them. Any suggestions or further information that is needed please let me know and TIA, BTW, haven't tried this on any of the RCs for 3.4 yet, have some work to do on the current version for work before I can upgrade my dev box, cheers, -- Paul Petersen -------------- next part -------------- Perl v5.8.5 under linux Apache::Session v1.6; Apache::Session::Generate::MD5 v2.1; Apache::Session::Lock::MySQL v1.00; Apache::Session::MySQL v1.01; Apache::Session::Serialize::Storable v1.00; Apache::Session::Store::DBI v1.02; Apache::Session::Store::MySQL v1.04; AutoLoader v5.60; base v2.06; Benchmark v1.06; bytes v1.01; Cache::Simple::TimedExpiry v0.21; Carp v1.03; CGI v3.05; CGI::Cookie v1.24; CGI::Fast v1.05; CGI::Util v1.5; Class::Container v0.11; Class::Data::Inheritable v0.02; Class::ReturnValue v0.52; constant v1.04; Cwd v2.19; Data::Dumper v2.121; DBD::mysql v2.1028; DBI v1.45; DBIx::SearchBuilder v1.11; Devel::StackTrace v1.11; Devel::StackTraceFrame v0.6; Digest::base v1.00; Digest::MD5 v2.33; DynaLoader v1.05; Encode v2.01; Encode::Alias v2.00; Encode::Config v2.00; Encode::Encoding v2.00; Errno v1.09; Exception::Class v1.19; Exception::Class::Base v1.2; Exporter v5.58; Exporter::Heavy v5.58; FCGI v0.67; Fcntl v1.05; fields v2.03; File::Basename v2.73; File::Glob v1.03; File::Path v1.06; File::Spec v0.87; File::Spec::Unix v1.5; File::Temp v0.14; FileHandle v2.01; HTML::Entities v1.27; HTML::Mason v1.26; HTML::Mason::CGIHandler v1.00; HTML::Mason::Exception v1.1; HTML::Mason::Exception::Abort v1.1; HTML::Mason::Exception::Compilation v1.1; HTML::Mason::Exception::Compilation::IncompatibleCompiler v1.1; HTML::Mason::Exception::Compiler v1.1; HTML::Mason::Exception::Decline v1.1; HTML::Mason::Exception::Params v1.1; HTML::Mason::Exception::Syntax v1.1; HTML::Mason::Exception::System v1.1; HTML::Mason::Exception::TopLevelNotFound v1.1; HTML::Mason::Exception::VirtualMethod v1.1; HTML::Mason::Exceptions v1.43; HTML::Parser v3.36; HTML::Scrubber v0.08; I18N::LangTags v0.35; integer v1.00; IO v1.21; IO::File v1.10; IO::Handle v1.24; IO::InnerFile v2.102 ; IO::Lines v2.103 ; IO::Scalar v2.105 ; IO::ScalarArray v2.103 ; IO::Seekable v1.09; IO::Wrap v2.102 ; IO::WrapTie v2.102 ; IPC::Open2 v1.01; IPC::Open3 v1.0106; lib v0.5565; List::Util v1.14; locale v1.00; Locale::Maketext v1.09; Locale::Maketext::Fuzzy v0.02; Locale::Maketext::Lexicon v0.44; Locale::Maketext::Lexicon::Gettext v0.12; Log::Dispatch v2.10; Log::Dispatch::Base v1.09; Log::Dispatch::Output v1.26; Log::Dispatch::Screen v1.17; Log::Dispatch::Syslog v1.18; Mail::Address v1.64; Mail::Field v1.64; Mail::Field::AddrList v1.64; Mail::Header v1.64; Mail::Internet v1.64; MIME::Base64 v3.05; MIME::Body v5.414; MIME::Decoder v5.414; MIME::Entity v5.414; MIME::Field::ContDisp v5.414; MIME::Field::ConTraEnc v5.414; MIME::Field::ContType v5.414; MIME::Field::ParamVal v5.414; MIME::Head v5.414; MIME::Parser v5.414; MIME::QuotedPrint v3.03; MIME::Tools v5.414; MIME::Words v5.414; Module::Versions::Report v1.02; overload v1.01; Params::Validate v0.74; POSIX v1.08; re v0.04; Regexp::Common v2.117; Regexp::Common::delimited v2.103; RT v3.2.2; Scalar::Util v1.14; SelectSaver v1.00; Socket v1.77; Storable v2.13; strict v1.03; Symbol v1.05; Sys::Hostname v1.11; Sys::Syslog v0.05; Text::Autoformat v1.12; Text::Quoted v1.8; Text::Reform v1.11; Text::Tabs v98.112801; Text::Template v1.44; Text::WikiFormat v0.72; Text::Wrapper v1.000; Time::HiRes v1.59; Time::JulianDay v2003.1125; Time::Local v1.1; Time::ParseDate v2003.1126; Time::Timezone v2003.0211; URI v1.34; URI::Escape v3.26; utf8 v1.04; vars v1.01; warnings v1.03; warnings::register v1.00; XSLoader v0.02; RT Variables RT::AmbiguousDayInPast 1 RT::BasePath /site/rt RT::CORE_CONFIG_FILE /site/rt/etc/RT_Config.pm RT::CanonicalizeEmailAddressMatch subdomain.example.com$ RT::CanonicalizeEmailAddressReplace example.com RT::CommentAddress RT_CommentAddressNotSet RT::CorrespondAddress RT_CorrespondAddressNotSet RT::DatabaseName rt3 RT::DatabasePassword Password not printed RT::DatabaseType mysql RT::DatabaseUser rt_user RT::DateDayBeforeMonth 1 RT::DefaultSearchResultFormat '__id__/TITLE:#', '__Subject__/TITLE:Subject', Status, QueueName, OwnerName, Priority, '__NEWLINE__', '', '__Requestors__', '__CreatedRelative__', '__ToldRelative__', '__LastUpdatedRelative__', '__TimeLeft__' RT::EmailOutputEncoding utf-8 RT::EtcPath /site/rt/etc RT::FriendlyFromLineFormat "%s via RT" <%s> RT::FriendlyToLineFormat "%s of Marchex Ticket #%s":; RT::LocalEtcPath /site/rt/local/etc RT::LocalLexiconPath /site/rt/local/po RT::LocalPath /site/rt/local RT::LogDir /site/rt/var/log RT::LogToFileNamed rt.log RT::LogToScreen error RT::LogToSyslog debug RT::LogoURL /NoAuth/images/rt.jpg RT::LoopsToRTOwner 1 RT::MailCommand sendmailpipe RT::MasonComponentRoot /site/rt/share/html RT::MasonDataDir /site/rt/var/mason_data RT::MasonLocalComponentRoot /site/rt/local/html RT::MasonSessionDir /site/rt/var/session_data RT::MaxAttachmentSize 10000000 RT::MaxInlineBody 13456 RT::MessageBoxWidth 72 RT::MessageBoxWrap HARD RT::MinimumPasswordLength Password not printed RT::MyRequestsLength 10 RT::MyTicketsLength 10 RT::Organization marchex.com RT::OwnerEmail monitoralert@marchex.com RT::RTAddressRegexp ^rt\@marchex.com$ RT::RecordOutgoingEmail 1 RT::SITE_CONFIG_FILE /site/rt/etc/RT_SiteConfig.pm RT::SendmailArguments -oi -t RT::SendmailPath /usr/sbin/sendmail RT::Timezone US/Pacific RT::UseFriendlyFromLine 1 RT::VERSION 3.2.2 RT::VarPath /site/rt/var RT::WebBaseURL http://rt.marchex.com RT::WebImagesURL /NoAuth/images/ RT::WebURL http://rt.marchex.com/ RT::rtname Marchex Perl configuration Summary of my perl5 (revision 5 version 8 subversion 5) configuration: Platform: osname=linux, osvers=, archname=i686-linux-thread-multi uname='linux laster 2.4.20-1-686 #1 sat mar 22 13:16:21 est 2003 i686 gnulinux ' config_args='-f config.sh' hint=previous, useposix=true, d_sigaction=define usethreads=define use5005threads=undef useithreads=define usemultiplicity=define useperlio=define d_sfio=undef uselargefiles=define usesocks=undef use64bitint=undef use64bitall=undef uselongdouble=undef usemymalloc=n, bincompat5005=undef Compiler: cc='cc', ccflags ='-D_REENTRANT -D_GNU_SOURCE -fno-strict-aliasing -D_LARGEFILE_SOURCE -D_FILE_OFFSET_BITS=64', optimize='-O3', cppflags='-D_REENTRANT -D_GNU_SOURCE -fno-strict-aliasing -D_REENTRANT -D_GNU_SOURCE -fno-strict-aliasing -D_LARGEFILE_SOURCE -D_FILE_OFFSET_BITS=64 -D_REENTRANT -D_GNU_SOURCE -fno-strict-aliasing -D_LARGEFILE_SOURCE -D_FILE_OFFSET_BITS=64 -D_REENTRANT -D_GNU_SOURCE -fno-strict-aliasing -D_LARGEFILE_SOURCE -D_FILE_OFFSET_BITS=64 -D_REENTRANT -D_GNU_SOURCE -fno-strict-aliasing -D_LARGEFILE_SOURCE -D_FILE_OFFSET_BITS=64 -D_REENTRANT -D_GNU_SOURCE -fno-strict-aliasing -D_LARGEFILE_SOURCE -D_FILE_OFFSET_BITS=64 -D_REENTRANT -D_GNU_SOURCE -fno-strict-aliasing -D_LARGEFILE_SOURCE -D_FILE_OFFSET_BITS=64 -D_REENTRANT -D_GNU_SOURCE -fno-strict-aliasing -D_LARGEFILE_SOURCE -D_FILE_OFFSET_BITS=64 -D_REENTRANT -D_GNU_SOURCE -fno-strict-aliasing -D_LARGEFILE_SOURCE -D_FILE_OFFSET_BITS=64 -D_REENTRANT -D_GNU_SOURCE -fno-strict-aliasing -D_LARGEFILE_SOURCE -D_FILE_OFFSET_BITS=64 -D_REENTRANT -D_GNU_SOURCE -fno-strict-aliasing -D_LARGEFILE_SOURCE -D_FILE_OFFSET_BITS=64 -D_REENTRANT -D_GNU_SOURCE -fno-strict-aliasing -D_LARGEFILE_SOURCE -D_FILE_OFFSET_BITS=64 -D_REENTRANT -D_GNU_SOURCE -fno-strict-aliasing -D_LARGEFILE_SOURCE -D_FILE_OFFSET_BITS=64 -D_REENTRANT -D_GNU_SOURCE -fno-strict-aliasing -D_LARGEFILE_SOURCE -D_FILE_OFFSET_BITS=64' ccversion='', gccversion='2.95.4 20011002 (Debian prerelease)', gccosandvers='' intsize=4, longsize=4, ptrsize=4, doublesize=8, byteorder=1234 d_longlong=define, longlongsize=8, d_longdbl=define, longdblsize=12 ivtype='long', ivsize=4, nvtype='double', nvsize=8, Off_t='off_t', lseeksize=8 alignbytes=4, prototype=define Linker and Libraries: ld='cc', ldflags =' -L/usr/local/lib' libpth=/usr/local/lib /lib /usr/lib libs=-lnsl -ldl -lm -lpthread -lc -lcrypt -lutil perllibs=-lnsl -ldl -lm -lpthread -lc -lcrypt -lutil libc=/lib/libc-2.3.2.so, so=so, useshrplib=false, libperl=libperl.a gnulibc_version='2.2.5' Dynamic Linking: dlsrc=dl_dlopen.xs, dlext=so, d_dlsymun=undef, ccdlflags='-rdynamic' cccdlflags='-fpic', lddlflags='-shared -L/usr/local/lib' ?|? RT 3.2.2 Copyright 1996-2004 From aharrison at gmail.com Thu Jan 13 14:20:35 2005 From: aharrison at gmail.com (Andy Harrison) Date: Thu Jan 13 14:23:28 2005 Subject: [Rt-devel] Content type, in the history In-Reply-To: References: Message-ID: On Thu, 13 Jan 2005 13:49:27 -0500, Patrick Rousseau-Mathieu wrote: > I want to know if someone know where's the content type of the email receive > via rt-mailgate is set (in the code) when we want to > watch the message in the history of the create ticket. > > I'd like to put this to: text/html and not text/plain. AFAIK it's done in rt3/lib/RT/Transaction_Overlay.pm, but honestly, I suspect that no good could come of this. Remember, anything that really is plain text will be rendered as html. So an e-mail like: > The five boxing wizards jump too quickly. > > 1 foo > 2 bar, 123 > 3 baz, 456 ...would be rendered as > The five boxing wizards jump too quickly. 1 foo 2 bar, 123 3 baz, 456 -- Andy Harrison From Ruslan.Zakirov at acronis.com Thu Jan 13 14:38:49 2005 From: Ruslan.Zakirov at acronis.com (Ruslan U. Zakirov) Date: Thu Jan 13 14:41:48 2005 Subject: [Rt-devel] Content type, in the history In-Reply-To: References: Message-ID: <41E6CE49.4040106@acronis.com> Patrick Rousseau-Mathieu wrote: > I want to know if someone know where's the content type of the email receive > via rt-mailgate is set (in the code) when we want to > watch the message in the history of the create ticket. Ticket history is html page, result of the HTTP request. HTTP response has own content-type and it's text/html. MIME message(email) is hierarchic tree of parts. Parts of the mail message has own content type. In tickets history you can see them inlined(text in the page) or not(via Download button). RT 3.0.x can inline only text/plain parts. Patches and solutions exists that allow to show text/html parts. RT 3.2.x inlines text/html by default. But if mail part content type is text/html then it's not safe to inline original html into history page, because of security reason: Cross Site Scripting. RT converts original text/html into text/plain with HTML::Scrubber module and display it as text block. See: html/Ticket/Elements/ShowTransactionAttachments html/Elements/ScrubHTML > > I'd like to put this to: text/html and not text/plain. > > Tnx > _______________________________________________ > Rt-devel mailing list > Rt-devel@lists.bestpractical.com > http://lists.bestpractical.com/cgi-bin/mailman/listinfo/rt-devel From jesse at bestpractical.com Thu Jan 13 15:47:43 2005 From: jesse at bestpractical.com (Jesse Vincent) Date: Thu Jan 13 15:47:56 2005 Subject: [Rt-devel] Requestor OR email like query from SearchBuilder sticks in mysql- RT v3.2.2 In-Reply-To: <1105643287.3037.128.camel@blades.marchex.com> References: <1105643287.3037.128.camel@blades.marchex.com> Message-ID: <20050113204743.GS24460@bestpractical.com> IIRC, we did some work on this for 3.2.3. Might be worth trying On Thu, Jan 13, 2005 at 11:08:07AM -0800, Paul Petersen wrote: > There is a query built through the SearchBuilder that consistently hangs > the query in mysql. I searched the mail list archives and the wiki and > only found similar situations that dealt with watchers and was fixed in > v3+ I'm hoping someone can reproduce to verify it is not an issue with > my setup and I apologize if I've missed prior reports of this. > > The particulars- > > RT v 3.2.2 > Mysql v 4.0.15 > Apache v 2.0.45 > Perl 5.8.5 > I've attached the config output from the RT System Configuration for any > other needed details. > > To reproduce- > Go to Tickets, New Query- Select OR radio button > Select Owner is (some user in system) > Select Requestor EmailAddress contains (part of email address of above > person) > > Variations of the above for the Requestor (RealName, is, etc.) produce > the same results. > > > The query that gets stuck on my system looks like this in the process > list: > | 355513 | rt_user | localhost | rt3 | Query | 63338 | > Sending data | SELECT COUNT(DISTINCT main.id) FROM Tickets main , Groups > Groups_1, CachedGroupMembers CachedGroupMembers_2, Users Users_3 WHERE > ((main.EffectiveId = main.id)) AND ((main.Status != 'deleted')) AND > ( ( (Users_3.EmailAddress LIKE '%ajcantu%')AND(Groups_1.Domain = > 'RT::Ticket-Role')AND(main.id = Groups_1.Instance)AND(Groups_1.Type = > 'Requestor')AND(Groups_1.id = > CachedGroupMembers_2.GroupId)AND(CachedGroupMembers_2.MemberId = > Users_3.id) ) OR(main.Owner = '414')) | > > The 63338 is the time it has been running and they never finish. I > found this after a user kept trying to do a search finding tickets he > owned and requested. There were so many of these that mysql finally > became unrepsonsive to RT and RT started to 500 waiting for it. I've > had to manually kill these processes to get rid of them. Any > suggestions or further information that is needed please let me know and > TIA, > > BTW, haven't tried this on any of the RCs for 3.4 yet, have some work to > do on the current version for work before I can upgrade my dev box, > cheers, > > -- > Paul Petersen > Perl v5.8.5 under linux > Apache::Session v1.6; > Apache::Session::Generate::MD5 v2.1; > Apache::Session::Lock::MySQL v1.00; > Apache::Session::MySQL v1.01; > Apache::Session::Serialize::Storable v1.00; > Apache::Session::Store::DBI v1.02; > Apache::Session::Store::MySQL v1.04; > AutoLoader v5.60; > base v2.06; > Benchmark v1.06; > bytes v1.01; > Cache::Simple::TimedExpiry v0.21; > Carp v1.03; > CGI v3.05; > CGI::Cookie v1.24; > CGI::Fast v1.05; > CGI::Util v1.5; > Class::Container v0.11; > Class::Data::Inheritable v0.02; > Class::ReturnValue v0.52; > constant v1.04; > Cwd v2.19; > Data::Dumper v2.121; > DBD::mysql v2.1028; > DBI v1.45; > DBIx::SearchBuilder v1.11; > Devel::StackTrace v1.11; > Devel::StackTraceFrame v0.6; > Digest::base v1.00; > Digest::MD5 v2.33; > DynaLoader v1.05; > Encode v2.01; > Encode::Alias v2.00; > Encode::Config v2.00; > Encode::Encoding v2.00; > Errno v1.09; > Exception::Class v1.19; > Exception::Class::Base v1.2; > Exporter v5.58; > Exporter::Heavy v5.58; > FCGI v0.67; > Fcntl v1.05; > fields v2.03; > File::Basename v2.73; > File::Glob v1.03; > File::Path v1.06; > File::Spec v0.87; > File::Spec::Unix v1.5; > File::Temp v0.14; > FileHandle v2.01; > HTML::Entities v1.27; > HTML::Mason v1.26; > HTML::Mason::CGIHandler v1.00; > HTML::Mason::Exception v1.1; > HTML::Mason::Exception::Abort v1.1; > HTML::Mason::Exception::Compilation v1.1; > HTML::Mason::Exception::Compilation::IncompatibleCompiler v1.1; > HTML::Mason::Exception::Compiler v1.1; > HTML::Mason::Exception::Decline v1.1; > HTML::Mason::Exception::Params v1.1; > HTML::Mason::Exception::Syntax v1.1; > HTML::Mason::Exception::System v1.1; > HTML::Mason::Exception::TopLevelNotFound v1.1; > HTML::Mason::Exception::VirtualMethod v1.1; > HTML::Mason::Exceptions v1.43; > HTML::Parser v3.36; > HTML::Scrubber v0.08; > I18N::LangTags v0.35; > integer v1.00; > IO v1.21; > IO::File v1.10; > IO::Handle v1.24; > IO::InnerFile v2.102 ; > IO::Lines v2.103 ; > IO::Scalar v2.105 ; > IO::ScalarArray v2.103 ; > IO::Seekable v1.09; > IO::Wrap v2.102 ; > IO::WrapTie v2.102 ; > IPC::Open2 v1.01; > IPC::Open3 v1.0106; > lib v0.5565; > List::Util v1.14; > locale v1.00; > Locale::Maketext v1.09; > Locale::Maketext::Fuzzy v0.02; > Locale::Maketext::Lexicon v0.44; > Locale::Maketext::Lexicon::Gettext v0.12; > Log::Dispatch v2.10; > Log::Dispatch::Base v1.09; > Log::Dispatch::Output v1.26; > Log::Dispatch::Screen v1.17; > Log::Dispatch::Syslog v1.18; > Mail::Address v1.64; > Mail::Field v1.64; > Mail::Field::AddrList v1.64; > Mail::Header v1.64; > Mail::Internet v1.64; > MIME::Base64 v3.05; > MIME::Body v5.414; > MIME::Decoder v5.414; > MIME::Entity v5.414; > MIME::Field::ContDisp v5.414; > MIME::Field::ConTraEnc v5.414; > MIME::Field::ContType v5.414; > MIME::Field::ParamVal v5.414; > MIME::Head v5.414; > MIME::Parser v5.414; > MIME::QuotedPrint v3.03; > MIME::Tools v5.414; > MIME::Words v5.414; > Module::Versions::Report v1.02; > overload v1.01; > Params::Validate v0.74; > POSIX v1.08; > re v0.04; > Regexp::Common v2.117; > Regexp::Common::delimited v2.103; > RT v3.2.2; > Scalar::Util v1.14; > SelectSaver v1.00; > Socket v1.77; > Storable v2.13; > strict v1.03; > Symbol v1.05; > Sys::Hostname v1.11; > Sys::Syslog v0.05; > Text::Autoformat v1.12; > Text::Quoted v1.8; > Text::Reform v1.11; > Text::Tabs v98.112801; > Text::Template v1.44; > Text::WikiFormat v0.72; > Text::Wrapper v1.000; > Time::HiRes v1.59; > Time::JulianDay v2003.1125; > Time::Local v1.1; > Time::ParseDate v2003.1126; > Time::Timezone v2003.0211; > URI v1.34; > URI::Escape v3.26; > utf8 v1.04; > vars v1.01; > warnings v1.03; > warnings::register v1.00; > XSLoader v0.02; > > > > RT Variables > RT::AmbiguousDayInPast 1 > RT::BasePath /site/rt > RT::CORE_CONFIG_FILE /site/rt/etc/RT_Config.pm > RT::CanonicalizeEmailAddressMatch subdomain.example.com$ > RT::CanonicalizeEmailAddressReplace example.com > RT::CommentAddress RT_CommentAddressNotSet > RT::CorrespondAddress RT_CorrespondAddressNotSet > RT::DatabaseName rt3 > RT::DatabasePassword Password not printed > RT::DatabaseType mysql > RT::DatabaseUser rt_user > RT::DateDayBeforeMonth 1 > RT::DefaultSearchResultFormat '__id__/TITLE:#', '__Subject__/TITLE:Subject', Status, QueueName, OwnerName, Priority, '__NEWLINE__', '', '__Requestors__', '__CreatedRelative__', '__ToldRelative__', '__LastUpdatedRelative__', '__TimeLeft__' > RT::EmailOutputEncoding utf-8 > RT::EtcPath /site/rt/etc > RT::FriendlyFromLineFormat "%s via RT" <%s> > RT::FriendlyToLineFormat "%s of Marchex Ticket #%s":; > RT::LocalEtcPath /site/rt/local/etc > RT::LocalLexiconPath /site/rt/local/po > RT::LocalPath /site/rt/local > RT::LogDir /site/rt/var/log > RT::LogToFileNamed rt.log > RT::LogToScreen error > RT::LogToSyslog debug > RT::LogoURL /NoAuth/images/rt.jpg > RT::LoopsToRTOwner 1 > RT::MailCommand sendmailpipe > RT::MasonComponentRoot /site/rt/share/html > RT::MasonDataDir /site/rt/var/mason_data > RT::MasonLocalComponentRoot /site/rt/local/html > RT::MasonSessionDir /site/rt/var/session_data > RT::MaxAttachmentSize 10000000 > RT::MaxInlineBody 13456 > RT::MessageBoxWidth 72 > RT::MessageBoxWrap HARD > RT::MinimumPasswordLength Password not printed > RT::MyRequestsLength 10 > RT::MyTicketsLength 10 > RT::Organization marchex.com > RT::OwnerEmail monitoralert@marchex.com > RT::RTAddressRegexp ^rt\@marchex.com$ > RT::RecordOutgoingEmail 1 > RT::SITE_CONFIG_FILE /site/rt/etc/RT_SiteConfig.pm > RT::SendmailArguments -oi -t > RT::SendmailPath /usr/sbin/sendmail > RT::Timezone US/Pacific > RT::UseFriendlyFromLine 1 > RT::VERSION 3.2.2 > RT::VarPath /site/rt/var > RT::WebBaseURL http://rt.marchex.com > RT::WebImagesURL /NoAuth/images/ > RT::WebURL http://rt.marchex.com/ > RT::rtname Marchex > Perl configuration > > Summary of my perl5 (revision 5 version 8 subversion 5) configuration: > Platform: > osname=linux, osvers=, archname=i686-linux-thread-multi > uname='linux laster 2.4.20-1-686 #1 sat mar 22 13:16:21 est 2003 i686 gnulinux ' > config_args='-f config.sh' > hint=previous, useposix=true, d_sigaction=define > usethreads=define use5005threads=undef useithreads=define usemultiplicity=define > useperlio=define d_sfio=undef uselargefiles=define usesocks=undef > use64bitint=undef use64bitall=undef uselongdouble=undef > usemymalloc=n, bincompat5005=undef > Compiler: > cc='cc', ccflags ='-D_REENTRANT -D_GNU_SOURCE -fno-strict-aliasing -D_LARGEFILE_SOURCE -D_FILE_OFFSET_BITS=64', > optimize='-O3', > cppflags='-D_REENTRANT -D_GNU_SOURCE -fno-strict-aliasing -D_REENTRANT -D_GNU_SOURCE -fno-strict-aliasing -D_LARGEFILE_SOURCE -D_FILE_OFFSET_BITS=64 -D_REENTRANT -D_GNU_SOURCE -fno-strict-aliasing -D_LARGEFILE_SOURCE -D_FILE_OFFSET_BITS=64 -D_REENTRANT -D_GNU_SOURCE -fno-strict-aliasing -D_LARGEFILE_SOURCE -D_FILE_OFFSET_BITS=64 -D_REENTRANT -D_GNU_SOURCE -fno-strict-aliasing -D_LARGEFILE_SOURCE -D_FILE_OFFSET_BITS=64 -D_REENTRANT -D_GNU_SOURCE -fno-strict-aliasing -D_LARGEFILE_SOURCE -D_FILE_OFFSET_BITS=64 -D_REENTRANT -D_GNU_SOURCE -fno-strict-aliasing -D_LARGEFILE_SOURCE -D_FILE_OFFSET_BITS=64 -D_REENTRANT -D_GNU_SOURCE -fno-strict-aliasing -D_LARGEFILE_SOURCE -D_FILE_OFFSET_BITS=64 -D_REENTRANT -D_GNU_SOURCE -fno-strict-aliasing -D_LARGEFILE_SOURCE -D_FILE_OFFSET_BITS=64 -D_REENTRANT -D_GNU_SOURCE -fno-strict-aliasing -D_LARGEFILE_SOURCE -D_FILE_OFFSET_BITS=64 -D_REENTRANT -D_GNU_SOURCE -fno-strict-aliasing -D_LARGEFILE_SOURCE -D_FILE_OFFSET_BITS=64 -D_REENTRANT -D_GNU_SOURCE -fno-strict-aliasing -D_LARGEFILE_SOURCE -D_FILE_OFFSET_BITS=64 -D_REENTRANT -D_GNU_SOURCE -fno-strict-aliasing -D_LARGEFILE_SOURCE -D_FILE_OFFSET_BITS=64 -D_REENTRANT -D_GNU_SOURCE -fno-strict-aliasing -D_LARGEFILE_SOURCE -D_FILE_OFFSET_BITS=64' > ccversion='', gccversion='2.95.4 20011002 (Debian prerelease)', gccosandvers='' > intsize=4, longsize=4, ptrsize=4, doublesize=8, byteorder=1234 > d_longlong=define, longlongsize=8, d_longdbl=define, longdblsize=12 > ivtype='long', ivsize=4, nvtype='double', nvsize=8, Off_t='off_t', lseeksize=8 > alignbytes=4, prototype=define > Linker and Libraries: > ld='cc', ldflags =' -L/usr/local/lib' > libpth=/usr/local/lib /lib /usr/lib > libs=-lnsl -ldl -lm -lpthread -lc -lcrypt -lutil > perllibs=-lnsl -ldl -lm -lpthread -lc -lcrypt -lutil > libc=/lib/libc-2.3.2.so, so=so, useshrplib=false, libperl=libperl.a > gnulibc_version='2.2.5' > Dynamic Linking: > dlsrc=dl_dlopen.xs, dlext=so, d_dlsymun=undef, ccdlflags='-rdynamic' > cccdlflags='-fpic', lddlflags='-shared -L/usr/local/lib' > > > > ?|? RT 3.2.2 Copyright 1996-2004 > _______________________________________________ > Rt-devel mailing list > Rt-devel@lists.bestpractical.com > http://lists.bestpractical.com/cgi-bin/mailman/listinfo/rt-devel -- From jesse at bestpractical.com Thu Jan 13 23:38:14 2005 From: jesse at bestpractical.com (Jesse Vincent) Date: Thu Jan 13 23:38:22 2005 Subject: [Rt-devel] RT 3.4 / Postgres / Search UI performance Message-ID: <20050114043814.GD24460@bestpractical.com> I've spent the evening doing some stress testing / performance work on RT 3.4's search page under postgres. I have a proposed patch, but I really need folks to beat on it to see if it improves things significantly and whether it seems to actually do the right thing. It's attached to this message. If folks see a speed boost without errors, it'll make it into 3.4.0. But this means taht I need sites running postgres (and other databases) to test it out. It's likey that the issue in question only manifests with a large number of queues. === html/Elements/SelectOwner ================================================================== --- html/Elements/SelectOwner (revision 2678) +++ html/Elements/SelectOwner (local) @@ -67,25 +67,19 @@ @objects = keys %{$cfqueues}; } else { - my $Queues = RT::Queues->new($session{CurrentUser}); - $Queues->UnLimit(); - while (my $Queue = $Queues->Next()) { - push( @objects, $Queue ); - } + # Let's check rights on an empty queue object. that will do a search for any queue. + my $queue = RT::Queue->new($session{'CurrentUser'}); + push( @objects, $queue ); } my %user_uniq_hash; foreach my $object (@objects) { my $Users = RT::Users->new($session{CurrentUser}); - $Users->WhoHaveRight(Right => 'OwnTicket', - Object => $object, - IncludeSystemRights => 1, - IncludeSuperusers => 0); - while (my $User = $Users->Next()) { + $Users->WhoHaveRight(Right => 'OwnTicket', Object => $object, IncludeSystemRights => 1, IncludeSuperusers => 0); while (my $User = $Users->Next()) { $user_uniq_hash{$User->Id()} = $User; } } -@users = sort { uc($a->Name) cmp uc($b->Name) } values %user_uniq_hash; +@users = sort { uc($a->Name) cmp uc($b->Name) } values %user_uniq_hash; <%ARGS> === lib/RT/Users_Overlay.pm ================================================================== --- lib/RT/Users_Overlay.pm (revision 2678) +++ lib/RT/Users_Overlay.pm (local) @@ -239,6 +239,7 @@ or as members of groups +If passed a queue object, with no id, it will find users who have that right for _any_ queue @@ -246,38 +247,121 @@ sub WhoHaveRight { my $self = shift; - my %args = ( Right => undef, - Object => undef, - IncludeSystemRights => undef, - IncludeSuperusers => undef, - IncludeSubgroupMembers => 1, - @_ ); + my %args = ( + Right => undef, + Object => undef, + IncludeSystemRights => undef, + IncludeSuperusers => undef, + IncludeSubgroupMembers => 1, + @_ + ); - if (defined $args{'ObjectType'} || defined $args{'ObjectId'}) { - $RT::Logger->crit("$self WhoHaveRight called with the Obsolete ObjectId/ObjectType API"); - return(undef); + if ( defined $args{'ObjectType'} || defined $args{'ObjectId'} ) { + $RT::Logger->crit( "$self WhoHaveRight called with the Obsolete ObjectId/ObjectType API"); + return (undef); } - my @privgroups; - my $Groups = RT::Groups->new($RT::SystemUser); - $Groups->WithRight(Right=> $args{'Right'}, - Object => $args{'Object'}, - IncludeSystemRights => $args{'IncludeSystemRights'}, - IncludeSuperusers => $args{'IncludeSuperusers'}); - while (my $Group = $Groups->Next()) { - push @privgroups, $Group->Id(); - } + + # Find only members of groups that have the right. - if (@privgroups) { - $self->WhoBelongToGroups(Groups => \@privgroups, - IncludeSubgroupMembers => $args{'IncludeSubgroupMembers'}); + my $acl = $self->NewAlias('ACL'); + my $groups = $self->NewAlias('Groups'); + my $userprinc = $self->{'princalias'}; + +# The cachedgroupmembers table is used for unrolling group memberships to allow fast lookups +# if we bind to CachedGroupMembers, we'll find all members of groups recursively. +# if we don't we'll find only 'direct' members of the group in question + my $cgm; + + if ( $args{'IncludeSubgroupMembers'} ) { + $cgm = $self->NewAlias('CachedGroupMembers'); } else { - # We don't have any group that matches -- make it impossible. - $self->Limit( FIELD => 'Id', VALUE => 'IS', OPERATOR => 'NULL' ); + $cgm = $self->NewAlias('GroupMembers'); } + +#Tie the users we're returning ($userprinc) to the groups that have rights granted to them ($groupprinc) + $self->Join( + ALIAS1 => $cgm, + FIELD1 => 'MemberId', + ALIAS2 => $userprinc, + FIELD2 => 'id' + ); + + $self->Join( + ALIAS1 => $groups, + FIELD1 => 'id', + ALIAS2 => $cgm, + FIELD2 => 'GroupId' + ); + +# {{{ Find only rows where the right granted is the one we're looking up or _possibly_ superuser + $self->Limit( + ALIAS => $acl, + FIELD => 'RightName', + OPERATOR => ( $args{Right} ? '=' : 'IS NOT' ), + VALUE => $args{Right} || 'NULL', + ENTRYAGGREGATOR => 'OR' + ); + + if ( $args{'IncludeSuperusers'} and $args{'Right'} ) { + $self->Limit( + ALIAS => $acl, + FIELD => 'RightName', + OPERATOR => '=', + VALUE => 'SuperUser', + ENTRYAGGREGATOR => 'OR' + ); + } + + # }}} + + my ( $or_check_ticket_roles, $or_check_roles ); + my $which_object = "$acl.ObjectType = 'RT::System'"; + + if ( defined $args{'Object'} ) { + if ( ref( $args{'Object'} ) eq 'RT::Ticket' ) { + $or_check_ticket_roles = " OR ( $groups.Domain = 'RT::Ticket-Role' AND $groups.Instance = " . $args{'Object'}->Id . ") "; + +# If we're looking at ticket rights, we also want to look at the associated queue rights. +# this is a little bit hacky, but basically, now that we've done the ticket roles magic, +# we load the queue object and ask all the rest of our questions about the queue. + $args{'Object'} = $args{'Object'}->QueueObj; + } + + # TODO XXX This really wants some refactoring + if ( ref( $args{'Object'} ) eq 'RT::Queue' ) { + $or_check_roles = " OR ( ( ($groups.Domain = 'RT::Queue-Role' "; + $or_check_roles .= "AND $groups.Instance = " . $args{'Object'}->id if ( $args{'Object'}->id ); + $or_check_roles .= ") $or_check_ticket_roles ) " . " AND $groups.Type = $acl.PrincipalType) "; + } + if ( $args{'IncludeSystemRights'} ) { + $which_object .= ' OR '; + } + else { + $which_object = ''; + } + $which_object .= " ($acl.ObjectType = '" . ref( $args{'Object'} ) . "'"; + if ( $args{'Object'}->id ) { + $which_object .= " AND $acl.ObjectId = " . $args{'Object'}->id; + } + + $which_object .= ") "; + } + $self->_AddSubClause( "WhichObject", "($which_object)" ); + $self->_AddSubClause( + "WhichGroup", + qq{ ( ( $acl.PrincipalId = $groups.id AND $acl.PrincipalType = 'Group' + AND ( $groups.Domain = 'SystemInternal' OR $groups.Domain = 'UserDefined' OR $groups.Domain = 'ACLEquivalence')) + $or_check_roles) } + ); + # only include regular RT users + $self->LimitToEnabled; + + # no system user + $self->Limit( ALIAS => $userprinc, FIELD => 'id', OPERATOR => '!=', VALUE => $RT::SystemUser->id); + } - # }}} # {{{ WhoBelongToGroups @@ -310,20 +394,14 @@ $cgm = $self->NewAlias('GroupMembers'); } - # {{{ Tie the users we're returning ($userprinc) to the groups that have rights granted to them ($groupprinc) + #Tie the users we're returning ($userprinc) to the groups that have rights granted to them ($groupprinc) $self->Join( ALIAS1 => $cgm, FIELD1 => 'MemberId', ALIAS2 => $userprinc, FIELD2 => 'id' ); - # }}} - # my $and_check_groups = "($cgm.GroupId = NULL"; foreach my $groupid (@{$args{'Groups'}}) { $self->Limit(ALIAS => $cgm, FIELD => 'GroupId', VALUE => $groupid, QUOTEVALUE => 0, ENTRYAGGREGATOR=> 'OR') - #$and_check_groups .= " OR $cgm.GroupId = $groupid"; } - #$and_check_groups .= ")"; - - #$self->_AddSubClause("WhichGroup", $and_check_groups); } # }}} -- From ajs at crankycanuck.ca Thu Jan 13 23:45:02 2005 From: ajs at crankycanuck.ca (Andrew Sullivan) Date: Thu Jan 13 23:48:00 2005 Subject: [Rt-devel] RT 3.4 / Postgres / Search UI performance In-Reply-To: <20050114043814.GD24460@bestpractical.com> References: <20050114043814.GD24460@bestpractical.com> Message-ID: <20050114044502.GB2408@phlogiston.dyndns.org> On Thu, Jan 13, 2005 at 11:38:14PM -0500, Jesse Vincent wrote: > It's attached to this message. If folks see a speed boost without > errors, it'll make it into 3.4.0. But this means taht I need sites > running postgres (and other databases) to test it out. It's likey that > the issue in question only manifests with a large number of queues. I will make sure we get this tested, but it'll take a few days. A -- Andrew Sullivan | ajs@crankycanuck.ca I remember when computers were frustrating because they *did* exactly what you told them to. That actually seems sort of quaint now. --J.D. Baldwin From jesse at bestpractical.com Thu Jan 13 23:53:10 2005 From: jesse at bestpractical.com (Jesse Vincent) Date: Thu Jan 13 23:53:14 2005 Subject: [Rt-devel] RT 3.4 / Postgres / Search UI performance In-Reply-To: <20050114044502.GB2408@phlogiston.dyndns.org> References: <20050114043814.GD24460@bestpractical.com> <20050114044502.GB2408@phlogiston.dyndns.org> Message-ID: <20050114045309.GE24460@bestpractical.com> On Thu, Jan 13, 2005 at 11:45:02PM -0500, Andrew Sullivan wrote: > On Thu, Jan 13, 2005 at 11:38:14PM -0500, Jesse Vincent wrote: > > It's attached to this message. If folks see a speed boost without > > errors, it'll make it into 3.4.0. But this means taht I need sites > > running postgres (and other databases) to test it out. It's likey that > > the issue in question only manifests with a large number of queues. > > I will make sure we get this tested, but it'll take a few days. *nod* I spent some time before dinner prodding at things and it seemed like 90% of the time was being spent in the selectowner widget. It was doing this loop: For each queue: Find each group that can own tickets in that queue. For each group: find its members. That's what we get for using a nice generalized layered API. The new call builds a single query that say: Find all the people who are members of groups that can own tickets an any queue. It's a bit hacky as yet, but hopefully it should work right. > A > > -- > Andrew Sullivan | ajs@crankycanuck.ca > I remember when computers were frustrating because they *did* exactly what > you told them to. That actually seems sort of quaint now. > --J.D. Baldwin > _______________________________________________ > Rt-devel mailing list > Rt-devel@lists.bestpractical.com > http://lists.bestpractical.com/cgi-bin/mailman/listinfo/rt-devel > -- From prouss at cae.com Fri Jan 14 11:01:19 2005 From: prouss at cae.com (Patrick Rousseau-Mathieu) Date: Fri Jan 14 11:04:33 2005 Subject: [Rt-devel] RTx::Atom download Message-ID: Hi, I just want to know if there is a place I can get RTx::Atom to install it and working on. Tnx Pat From dkrt at navahonetworks.com Fri Jan 14 11:22:56 2005 From: dkrt at navahonetworks.com (David Kerry) Date: Fri Jan 14 11:25:59 2005 Subject: [Rt-devel] RT 3.4 / Postgres / Search UI performance In-Reply-To: <20050114043814.GD24460@bestpractical.com> References: <20050114043814.GD24460@bestpractical.com> Message-ID: <20050114162256.GA15676@navahonetworks.com> On Thu, Jan 13, 2005 at 11:38:14PM -0500, Jesse Vincent wrote: > > I've spent the evening doing some stress testing / performance work on > RT 3.4's search page under postgres. I have a proposed patch, but I > really need folks to beat on it to see if it improves things > significantly and whether it seems to actually do the right thing. > > It's attached to this message. If folks see a speed boost without > errors, it'll make it into 3.4.0. But this means taht I need sites > running postgres (and other databases) to test it out. It's likey that > the issue in question only manifests with a large number of queues. Wow... huge improvement here. 20+ seconds to bring up this one page is now down to around 2 seconds. This patch gets my vote of approval :-) For what it's worth, we have 18 queues defined, which explains why the situation was so bad for us. -- David Kerry From jesse at bestpractical.com Fri Jan 14 13:52:50 2005 From: jesse at bestpractical.com (Jesse Vincent) Date: Fri Jan 14 13:52:57 2005 Subject: [Rt-devel] RT 3.4 / Postgres / Search UI performance In-Reply-To: <20050114162256.GA15676@navahonetworks.com> References: <20050114043814.GD24460@bestpractical.com> <20050114162256.GA15676@navahonetworks.com> Message-ID: <20050114185250.GB24460@bestpractical.com> On Fri, Jan 14, 2005 at 11:22:56AM -0500, David Kerry wrote: > On Thu, Jan 13, 2005 at 11:38:14PM -0500, Jesse Vincent wrote: > > > > I've spent the evening doing some stress testing / performance work on > > RT 3.4's search page under postgres. I have a proposed patch, but I > > really need folks to beat on it to see if it improves things > > significantly and whether it seems to actually do the right thing. > > > > It's attached to this message. If folks see a speed boost without > > errors, it'll make it into 3.4.0. But this means taht I need sites > > running postgres (and other databases) to test it out. It's likey that > > the issue in question only manifests with a large number of queues. > > Wow... huge improvement here. 20+ seconds to bring up this > one page is now down to around 2 seconds. Is the list of possible owners _correct_? J > > This patch gets my vote of approval :-) > > For what it's worth, we have 18 queues defined, which explains why > the situation was so bad for us. > > -- > David Kerry > -- From dkrt at navahonetworks.com Fri Jan 14 14:09:21 2005 From: dkrt at navahonetworks.com (David Kerry) Date: Fri Jan 14 14:18:41 2005 Subject: [Rt-devel] RT 3.4 / Postgres / Search UI performance In-Reply-To: <20050114185250.GB24460@bestpractical.com> References: <20050114043814.GD24460@bestpractical.com> <20050114162256.GA15676@navahonetworks.com> <20050114185250.GB24460@bestpractical.com> Message-ID: <20050114190921.GB16247@navahonetworks.com> On Fri, Jan 14, 2005 at 01:52:50PM -0500, Jesse Vincent wrote: > > On Fri, Jan 14, 2005 at 11:22:56AM -0500, David Kerry wrote: > > On Thu, Jan 13, 2005 at 11:38:14PM -0500, Jesse Vincent wrote: > > > > > > I've spent the evening doing some stress testing / performance work on > > > RT 3.4's search page under postgres. I have a proposed patch, but I > > > really need folks to beat on it to see if it improves things > > > significantly and whether it seems to actually do the right thing. > > > > > > It's attached to this message. If folks see a speed boost without > > > errors, it'll make it into 3.4.0. But this means taht I need sites > > > running postgres (and other databases) to test it out. It's likey that > > > the issue in question only manifests with a large number of queues. > > > > Wow... huge improvement here. 20+ seconds to bring up this > > one page is now down to around 2 seconds. > > Is the list of possible owners _correct_? > > J It appears to be correct. -- David Kerry From dkrt at navahonetworks.com Fri Jan 14 14:14:31 2005 From: dkrt at navahonetworks.com (David Kerry) Date: Fri Jan 14 14:19:45 2005 Subject: [Rt-devel] Constantly updating attachments? Message-ID: <20050114191431.GC16247@navahonetworks.com> I've been looking at the sql queries RT has been using to see if some operations can be sped up... anyway, I discovered this one that popped out when viewing an individual ticket: SELECT content FROM Attachments WHERE id = '279982'; followed immediately after by this: UPDATE Attachments SET Headers= '......' WHERE id= '279982'; This seems a little odd - RT updates the attachments on a ticket even when I'm just viewing it? -- David Kerry From jesse at bestpractical.com Fri Jan 14 14:21:24 2005 From: jesse at bestpractical.com (Jesse Vincent) Date: Fri Jan 14 14:21:30 2005 Subject: [Rt-devel] Constantly updating attachments? In-Reply-To: <20050114191431.GC16247@navahonetworks.com> References: <20050114191431.GC16247@navahonetworks.com> Message-ID: <20050114192124.GE24460@bestpractical.com> On Fri, Jan 14, 2005 at 02:14:31PM -0500, David Kerry wrote: > > I've been looking at the sql queries RT has been using to see if some > operations can be sped up... anyway, I discovered this one that > popped out when viewing an individual ticket: > > SELECT content FROM Attachments WHERE id = '279982'; > > followed immediately after by this: > > UPDATE Attachments SET Headers= '......' WHERE id= '279982'; > > > This seems a little odd - RT updates the attachments on a ticket even > when I'm just viewing it? If it sees at an attachment that doesn't have a content-length header, it fixes it. Are you seeing it update the same attachment multiple times? (The content-length header is an important performance thing that wasn't always there) > > -- > David Kerry > > > _______________________________________________ > Rt-devel mailing list > Rt-devel@lists.bestpractical.com > http://lists.bestpractical.com/cgi-bin/mailman/listinfo/rt-devel > -- From dkrt at navahonetworks.com Fri Jan 14 14:34:23 2005 From: dkrt at navahonetworks.com (David Kerry) Date: Fri Jan 14 14:37:26 2005 Subject: [Rt-devel] Constantly updating attachments? In-Reply-To: <20050114192124.GE24460@bestpractical.com> References: <20050114191431.GC16247@navahonetworks.com> <20050114192124.GE24460@bestpractical.com> Message-ID: <20050114193423.GD16247@navahonetworks.com> On Fri, Jan 14, 2005 at 02:21:24PM -0500, Jesse Vincent wrote: > On Fri, Jan 14, 2005 at 02:14:31PM -0500, David Kerry wrote: > > > > I've been looking at the sql queries RT has been using to see if some > > operations can be sped up... anyway, I discovered this one that > > popped out when viewing an individual ticket: > > > > SELECT content FROM Attachments WHERE id = '279982'; > > > > followed immediately after by this: > > > > UPDATE Attachments SET Headers= '......' WHERE id= '279982'; > > > > > > This seems a little odd - RT updates the attachments on a ticket even > > when I'm just viewing it? > > If it sees at an attachment that doesn't have a content-length header, > it fixes it. Are you seeing it update the same attachment multiple > times? (The content-length header is an important performance thing that > wasn't always there) Ahh... no - this is only happening once per ticket, as intended then. -- David Kerry From lucas at dextra.com.br Mon Jan 17 05:35:26 2005 From: lucas at dextra.com.br (Lucas Fernando Rosada) Date: Mon Jan 17 05:45:29 2005 Subject: [Rt-devel] Extending RT: adding columns to the database Message-ID: <20050117103526.17323.qmail@hm28.locaweb.com.br> Good morning for all. I'm trying to add some more types of custom fields input data, like masked fields, external database feed fields, etc. But to develop this, the RT database structure is not enough, so I needed to add some more columns to the CustomFields table, then I extended the source code of RT but I'm having a big issue (please correct me if I'm wrong): I read that the RT perl library, like CustomField.pm, Ticket.pm, etc. are automatically generated by DBIx::Record, just after the restart processes of the webserver, so I don't have to maintain or add code into them. But it isn't being updated, I restart the webserver and it doesn't work. Cause I'm in a hurry to test the new extension, I have added code manually to the CustomField.pm class, but I know that it isn't correct form and it can suddenly desapear, losing my changes. Oh! Only highlighting: I did make my changes at CustomFields_Vendor, at my local directory. I'd like to know how to make the RT system regenerate the CustomField.pm from the database structure. Regards, Lucas F. Rosada Dextra Sistemas lucas at dextra.com.br From prouss at cae.com Tue Jan 18 10:13:26 2005 From: prouss at cae.com (Patrick Rousseau-Mathieu) Date: Tue Jan 18 10:17:20 2005 Subject: [Rt-devel] Rt::Atom , client Message-ID: Hi, I'm trying to create a new ticket with atom (from php/java) and i just dont get it, i'd like to know if someone of you have an exemple of the http request header budy and content with the xml+atom so i can saw what it should be like. If not, well a url or something to refer to (with code or http request,header..etc). Thanks a lot Patrick. From lucas at dextra.com.br Tue Jan 18 13:02:53 2005 From: lucas at dextra.com.br (Lucas Fernando Rosada) Date: Tue Jan 18 13:06:29 2005 Subject: [Rt-devel] Updating RT classes Message-ID: <20050118180253.11183.qmail@hm28.locaweb.com.br> What do I need to do to RT update the Perl Classes, that do the database access? I didn't found it in any documentation. Thank you, Lucas. lucas@dextra.com.br From sturner at MIT.EDU Tue Jan 18 13:22:50 2005 From: sturner at MIT.EDU (Stephen Turner) Date: Tue Jan 18 13:26:38 2005 Subject: [Rt-devel] Updating RT classes In-Reply-To: <20050118180253.11183.qmail@hm28.locaweb.com.br> Message-ID: <5.2.1.1.2.20050118132152.01e3e468@po14.mit.edu> At Tuesday 1/18/2005 01:02 PM, Lucas Fernando Rosada wrote: >What do I need to do to RT update the Perl Classes, that do the database >access? I didn't found it in any documentation. > >Thank you, > >Lucas. >lucas@dextra.com.br Lucas, There's some good information on this at: http://wiki.bestpractical.com/index.cgi?CleanlyCustomizeRT Steve Stephen Turner Senior Programmer/Analyst - Client Support Services Information Services and Technology (IS&T) sturner@mit.edu From lucas at dextra.com.br Tue Jan 18 14:00:26 2005 From: lucas at dextra.com.br (Lucas Fernando Rosada) Date: Tue Jan 18 14:03:56 2005 Subject: [Rt-devel] Re: Updating RT classes In-Reply-To: <5.2.1.1.2.20050118132152.01e3e468@po14.mit.edu> References: <5.2.1.1.2.20050118132152.01e3e468@po14.mit.edu> Message-ID: <20050118190026.32724.qmail@hm28.locaweb.com.br> Stephen, in this case, the factory file (rtpath/sbin/factory) will be needed. It's this file that build the perl data access file from the database structure. Thank you, Lucas On Tue, 18 Jan 2005 13:22:50 -0500, Stephen Turner escreveu: > De: Stephen Turner > Data: Tue, 18 Jan 2005 13:22:50 -0500 > Para: "Lucas Fernando Rosada" , rt-devel@lists.bestpractical.com > Assunto: Re: [Rt-devel] Updating RT classes > > At Tuesday 1/18/2005 01:02 PM, Lucas Fernando Rosada wrote: > > >What do I need to do to RT update the Perl Classes, that do the database > >access? I didn't found it in any documentation. > > > >Thank you, > > > >Lucas. > >lucas@dextra.com.br > > Lucas, > > There's some good information on this at: > > http://wiki.bestpractical.com/index.cgi?CleanlyCustomizeRT > > Steve > > Stephen Turner > > Senior Programmer/Analyst - Client Support Services > Information Services and Technology (IS&T) > > sturner@mit.edu > > > > From yoavd at qballtech.net Wed Jan 19 03:43:15 2005 From: yoavd at qballtech.net (Yoav Daniely) Date: Wed Jan 19 04:17:37 2005 Subject: [Rt-devel] I would like to report a bug Message-ID: <20050119084320.BEB611276CE@hawk.qballtech.com> Greetings everybody, I'm testing RT 3.4.0RC5 , everything was very good for a while but after adding a few "User custom fields" I've been unable to see user details from the ticket view ("More about ...."). This does not happen with a user which has the "super user" permission. I tried granting the user with various permissions, that includes a specific permission for the custom fields and a global SeeCustomFields , AssignCustomFields and AdminCustomField. Is this a bug? if so I'm reporting it (hopefully this is the right spot :)) Regards, Yoav Those are the global permissions for the user: AdminCustomField AdminOwnPersonalGroups AdminUsers AssignCustomFields CreateSavedSearch DelegateRights EditSavedSearches LoadSavedSearch ModifyACL ModifyCustomField ModifySelf ModifyTicket OwnTicket SeeCustomField SeeGroup ShowACL ShowOutgoingEmail ShowSavedSearches Watch -------------- next part -------------- An HTML attachment was scrubbed... URL: http://bestpractical.com/pipermail/rt-devel/attachments/20050119/bdbb8e2a/attachment.html From aharrison at gmail.com Wed Jan 19 14:40:53 2005 From: aharrison at gmail.com (Andy Harrison) Date: Wed Jan 19 14:44:36 2005 Subject: [Rt-devel] RT Wiki: How to remove spam correctly Message-ID: To fix a spammed page, *DO NOT* edit the contents of the wiki page. I'm noticing folks are just editing the text and deleting the spam portions, then saving the page. Keep in mind that the spammers are often destroying some of the existing content of wiki pages. Simply wiping out the spam does not restore any of the deleted content. Instead, click "Revisions" and cycle through "Previous" revisions until you find the last good version without spam. Click Edit. Click Save. Done. If the spamming bastards have put content in that lobotomizes kwiki when it tries to display it, you need to construct your links manually. As Jesse previously instructed: http://wiki.bestpractical.com/index.cgi?page_name=HomePage&action=revisions&revision_id=-1 Give it the appropriate page_name. The -1 is the last revision before the current version, but you can also specify -2 to go the one before that since the previous version is often spammed as well. After you construct the link and find the last good revision, do the same thing. Edit. Save. Done. -- Andy Harrison From prouss at cae.com Thu Jan 20 12:07:33 2005 From: prouss at cae.com (Patrick Rousseau-Mathieu) Date: Thu Jan 20 12:12:27 2005 Subject: [Rt-devel] RT:Atom POST problem Message-ID: Hi, I've got a problem trying to create a ticket or a user with the POST statement of HTTP. Rt always telling me that a certain properties is not set. I've try this with to get info on Users with the GET statement and its working pretty well and giving me all the user of RT: GET /rt/Atom/0.3/RT-Users HTTP/1.1 \r\n Host: 127.0.0.1 \r\n Accept: text/xml,application/xml,application/xhtml+xml,text/html;q=0.9,text/plain;q= 0.8,image/png,*/*;q=0.5 \r\n Accept-Language: en-us,en;q=0.5 \r\n Accept-Encoding: gzip,deflate \r\n Accept-Charset: ISO-8859-1,utf-8;q=0.7,*;q=0.7 \r\n Keep-Alive: 300 \r\n Connection: keep-alive \r\n Cookie: RT_SID_rttest.cae.ca.80=9b2c09c36a9b0312a7cc9b9185c30fda \r\n \r\n (NOTE: \r\n are the "enter" that i put at the end of each line, in php") But with the POST one with the POST /rt/Atom/0.3/Users HTTP/1.1 it says: "Must specify 'Name' attribute". It look like rt dont parse the xml: POST /rt/Atom/0.3/Users HTTP/1.1 \r\n Host: 127.0.0.1 \r\n Keep-Alive: 300 \r\n Connection: keep-alive \r\n Cookie: RT_SID_rttest.cae.ca.80=9b2c09c36a9b0312a7cc9b9185c30fda \r\n Content-Type: text/xml,application/x.atom+xml,*/* \r\n r\n Patrickprouss@cae.com If someone can give me an exemple of a POST that work or something to help it will be verry appreciated. Thanks. Patrick R. M. From jesse at bestpractical.com Fri Jan 21 07:39:17 2005 From: jesse at bestpractical.com (Jesse Vincent) Date: Fri Jan 21 07:39:22 2005 Subject: [Rt-devel] Re: New czech translation - corrected errors In-Reply-To: References: Message-ID: <20050121123917.GN24460@bestpractical.com> On Fri, Jan 21, 2005 at 12:01:59PM +0100, Jan Okrouhly wrote: > Hi Jesse, > > the attachments contains some important corrections of translation of RT > 3.2.2 cs.po file. Thanks! Is this an update of the previous one you sent? I'd already integrated that into the 3.2.3 release candidate. I'll diff this one and add it in. Best, Jesse > > Greets > > Jan Okrouhl? > > //------------------------------------\-\-\+\-\-\---okrouhly@civ.zcu.cz-\\ > |Laboratory for Computer Science |don't call: +420 377 632 837 | > |University of West Bohemia | location: Univerzitni 22 | > |Americka 42, 306 14 Pilsen, Czech Republic| room: UI414 | > +---------------------------------------73-!-de-OK1INC@OK0PPL.#BOH.CZE.EU+ > | CESNET z.s.p.o., Zikova 4, Prague | phone: +420 224 355 263 | > > \\________________________________________|______room:_90______________// -- From okrouhly at civ.zcu.cz Fri Jan 21 07:42:41 2005 From: okrouhly at civ.zcu.cz (Jan Okrouhly) Date: Fri Jan 21 07:46:33 2005 Subject: [Rt-devel] Re: New czech translation - corrected errors In-Reply-To: <20050121123917.GN24460@bestpractical.com> References: <20050121123917.GN24460@bestpractical.com> Message-ID: On Fri, 21 Jan 2005, Jesse Vincent wrote: > > > > On Fri, Jan 21, 2005 at 12:01:59PM +0100, Jan Okrouhly wrote: >> Hi Jesse, >> >> the attachments contains some important corrections of translation of RT >> 3.2.2 cs.po file. > > Thanks! Is this an update of the previous one you sent? I'd already Yes it is, some users found that there was comma delimited by mistake translated to space delimited, what's terribly wrong. Nothing more changed. > integrated that into the 3.2.3 release candidate. I'll diff this one and > add it in. Thank You > Best, > Jesse > >> >> Greets >> >> Jan Okrouhl? >> >> //------------------------------------\-\-\+\-\-\---okrouhly@civ.zcu.cz-\\ >> |Laboratory for Computer Science |don't call: +420 377 632 837 | >> |University of West Bohemia | location: Univerzitni 22 | >> |Americka 42, 306 14 Pilsen, Czech Republic| room: UI414 | >> +---------------------------------------73-!-de-OK1INC@OK0PPL.#BOH.CZE.EU+ >> | CESNET z.s.p.o., Zikova 4, Prague | phone: +420 224 355 263 | >> >> \\________________________________________|______room:_90______________// > > > > -- > > Jan Okrouhl? //------------------------------------\-\-\+\-\-\---okrouhly@civ.zcu.cz-\\ |Laboratory for Computer Science |don't call: +420 377 632 837 | |University of West Bohemia | location: Univerzitni 22 | |Americka 42, 306 14 Pilsen, Czech Republic| room: UI414 | +---------------------------------------73-!-de-OK1INC@OK0PPL.#BOH.CZE.EU+ | CESNET z.s.p.o., Zikova 4, Prague | phone: +420 224 355 263 | \\________________________________________|______room:_90______________// From msf at kisoku.net Fri Jan 21 17:04:33 2005 From: msf at kisoku.net (Mathieu Sauve-Frankel) Date: Fri Jan 21 17:28:41 2005 Subject: [Rt-devel] 3.4.0rc5 CustomFields issue Message-ID: <20050121220433.GC11672@kisoku.net> If you define a Ticket CustomField and then enable it both globally and on a particular queue, that Customfield shows up twice on both the Create and Modify pages. -- Mathieu Sauve-Frankel msf@kisoku.net || msf@openbsd.org From rt at chaka.net Fri Jan 21 17:34:54 2005 From: rt at chaka.net (Todd Chapman) Date: Fri Jan 21 18:17:25 2005 Subject: [Rt-devel] Generating a stack trace? Message-ID: <20050121223454.GK27116@chaka.net> How do I generate a trace of all the RT/SB API calls that result from a click in the web interface? Thanks. -Todd From Ruslan.Zakirov at acronis.com Fri Jan 21 18:34:29 2005 From: Ruslan.Zakirov at acronis.com (Ruslan U. Zakirov) Date: Fri Jan 21 18:38:25 2005 Subject: [Rt-devel] Generating a stack trace? In-Reply-To: <20050121223454.GK27116@chaka.net> References: <20050121223454.GK27116@chaka.net> Message-ID: <41F19185.5000307@acronis.com> Todd Chapman wrote: > How do I generate a trace of all the RT/SB API calls that result > from a click in the web interface? Apache::DProf See http://wiki.bestpractical.com/index.cgi?ApacheConfig dprofpp -T tmon.out This will output call tree. Or use "require Carp; Carp::cluck();" if you want stack trace from particular place. > > Thanks. > > -Todd > _______________________________________________ > Rt-devel mailing list > Rt-devel@lists.bestpractical.com > http://lists.bestpractical.com/cgi-bin/mailman/listinfo/rt-devel From Ruslan.Zakirov at acronis.com Sat Jan 22 13:55:26 2005 From: Ruslan.Zakirov at acronis.com (Ruslan U. Zakirov) Date: Sat Jan 22 13:59:27 2005 Subject: [Rt-devel] [PATCH] Quote identifiers. Message-ID: <41F2A19E.3050808@acronis.com> Hello. Quote all identifiers with DBI. Best regards. Ruslan. -------------- next part -------------- A non-text attachment was scrubbed... Name: quote_identifier.patch Type: application/aegis-patch Size: 1338 bytes Desc: not available Url : http://bestpractical.com/pipermail/rt-devel/attachments/20050122/544fec84/quote_identifier.bin From jesse at bestpractical.com Sat Jan 22 14:01:26 2005 From: jesse at bestpractical.com (Jesse Vincent) Date: Sat Jan 22 14:01:30 2005 Subject: [Rt-devel] [PATCH] Quote identifiers. In-Reply-To: <41F2A19E.3050808@acronis.com> References: <41F2A19E.3050808@acronis.com> Message-ID: <20050122190125.GS24460@bestpractical.com> On Sat, Jan 22, 2005 at 09:55:26PM +0300, Ruslan U. Zakirov wrote: > Hello. > Quote all identifiers with DBI. > Best regards. Ruslan. Any reason you're assigning $self->dbi to $dbi and using that? > _______________________________________________ > Rt-devel mailing list > Rt-devel@lists.bestpractical.com > http://lists.bestpractical.com/cgi-bin/mailman/listinfo/rt-devel -- From Ruslan.Zakirov at acronis.com Sat Jan 22 14:04:01 2005 From: Ruslan.Zakirov at acronis.com (Ruslan U. Zakirov) Date: Sat Jan 22 14:08:03 2005 Subject: [Rt-devel] [PATCH] Quote identifiers. In-Reply-To: <20050122190125.GS24460@bestpractical.com> References: <41F2A19E.3050808@acronis.com> <20050122190125.GS24460@bestpractical.com> Message-ID: <41F2A3A1.1000701@acronis.com> Jesse Vincent wrote: > > > On Sat, Jan 22, 2005 at 09:55:26PM +0300, Ruslan U. Zakirov wrote: > >> Hello. >>Quote all identifiers with DBI. >> Best regards. Ruslan. > > > > Any reason you're assigning $self->dbi to $dbi and using that? you mean dbh? yes, it's required, because quote_identifier like quote is method of the database handle not statement. > > > >>_______________________________________________ >>Rt-devel mailing list >>Rt-devel@lists.bestpractical.com >>http://lists.bestpractical.com/cgi-bin/mailman/listinfo/rt-devel > > > From jesse at bestpractical.com Sat Jan 22 14:14:04 2005 From: jesse at bestpractical.com (Jesse Vincent) Date: Sat Jan 22 14:14:09 2005 Subject: [Rt-devel] [PATCH] Quote identifiers. In-Reply-To: <41F2A3A1.1000701@acronis.com> References: <41F2A19E.3050808@acronis.com> <20050122190125.GS24460@bestpractical.com> <41F2A3A1.1000701@acronis.com> Message-ID: <20050122191404.GT24460@bestpractical.com> > > > >Any reason you're assigning $self->dbi to $dbi and using that? > you mean dbh? yes, it's required, because quote_identifier like quote is > method of the database handle not statement. > Er. Sorry. There's code of the form: my $dbh = $self->dbh; $dbh->quote_identifier(); Since it's just a reference, I don't quite understand the point. From Ruslan.Zakirov at acronis.com Sat Jan 22 14:26:47 2005 From: Ruslan.Zakirov at acronis.com (Ruslan U. Zakirov) Date: Sat Jan 22 14:30:50 2005 Subject: [Rt-devel] [PATCH] Quote identifiers. In-Reply-To: <20050122191404.GT24460@bestpractical.com> References: <41F2A19E.3050808@acronis.com> <20050122190125.GS24460@bestpractical.com> <41F2A3A1.1000701@acronis.com> <20050122191404.GT24460@bestpractical.com> Message-ID: <41F2A8F7.9020907@acronis.com> Jesse Vincent wrote: >>>Any reason you're assigning $self->dbi to $dbi and using that? >> >>you mean dbh? yes, it's required, because quote_identifier like quote is >>method of the database handle not statement. >> > > > > Er. Sorry. There's code of the form: > > > my $dbh = $self->dbh; > > $dbh->quote_identifier(); > > > Since it's just a reference, I don't quite understand the point. Speed. See next benchmark #!/usr/bin/perl -w use strict; use Benchmark qw(cmpthese); my $obj = new Dummy; cmpthese (-6, { 'with $dbh' => sub { $obj->w_cache(); }, 'without $dbh' => sub { $obj->wo_cache() }, }); package Dummy; sub new { return bless { dbh => { foo => 'bar' } }, 'Dummy' } sub dbh { my $self = shift; return $self->{'dbh'} } sub w_cache { my $self = shift; my $dbh = $self->dbh; for(1..5) { my $x = $dbh->{foo} } } sub wo_cache { my $self = shift; for(1..5) {my $x = $self->dbh->{foo} } } Results: Rate without $dbh with $dbh without $dbh 43465/s -- -40% with $dbh 72127/s 66% -- Something similar happens in sub Insert I call quote_identifier in map context for each column, also dbh sub in SB::Handle more complex then in benchmark. Similar situation in UpdateRecordValue. From jesse at bestpractical.com Sat Jan 22 15:01:16 2005 From: jesse at bestpractical.com (Jesse Vincent) Date: Sat Jan 22 15:01:28 2005 Subject: [Rt-devel] RT 3.4.0rc6 Message-ID: <20050122200116.GV24460@bestpractical.com> RT 3.4.0rc6 is now available. It features a number of small cleanups and bugfixes and some important fixes to Custom Field searching. We also integrated an important performance improvement for the ticket search page. Combined with DBIx-SearchBuilder 1.21, this version of RT should work cleanly with DBD::Oracle 1.16. Jesse Changelog since rc5: ---------------------------------------------------------------------- r3926 (orig r2131): jesse | 2005-01-22 14:43:46 -0500 r3924@hualien: jesse | 2005-01-22T19:34:52.453359Z Bumped to 3.4.0rc6 ---------------------------------------------------------------------- r3914 (orig r2123): jesse | 2005-01-22 11:57:23 -0500 r3910@hualien: jesse | 2005-01-22T16:06:58.731752Z Custom field searches should only search on values that haven't been deleted ---------------------------------------------------------------------- r3913 (orig r2122): jesse | 2005-01-22 11:57:11 -0500 r3909@hualien: jesse | 2005-01-22T15:42:46.936767Z Bullet-proofing for custom fields with a set but limited number of values ---------------------------------------------------------------------- r3912 (orig r2121): jesse | 2005-01-22 11:57:01 -0500 r3907@hualien: jesse | 2005-01-22T14:48:14.100308Z TicketSQL search fixes, especially to Custom field operations ---------------------------------------------------------------------- r3911 (orig r2120): jesse | 2005-01-22 11:53:50 -0500 r3904@hualien: jesse | 2005-01-22T13:00:19.775653Z The column map callback was broken by performance work ---------------------------------------------------------------------- r3903 (orig r2119): jesse | 2005-01-22 06:26:14 -0500 r3902@hualien: jesse | 2005-01-22T11:21:43.212477Z RT-Ticket: 6366 RT-Update: correspond Cleaned up the upgrade procedure to give better messages on success. Added content so it won't bomb out on errors for the 3.4 upgrades. (We could have altered the script to not error out on missing content, but that would obscure errors where the user flubbed the path. ---------------------------------------------------------------------- r3896 (orig r2116): jesse | 2005-01-19 17:20:57 -0500 r3893@hualien: jesse | 2005-01-19T18:39:39.850646Z Updated version number used by development builds to note that this is the 3.4 series ---------------------------------------------------------------------- r3895 (orig r2115): jesse | 2005-01-19 17:20:48 -0500 r3892@hualien: jesse | 2005-01-19T18:38:38.233900Z Added better error checking for attachment insertion failure. (This unmasked more instances of oracle brokenness) ---------------------------------------------------------------------- r3894 (orig r2114): jesse | 2005-01-19 17:20:39 -0500 r3891@hualien: jesse | 2005-01-19T18:31:15.604733Z Minor formatting cleanups to tests ---------------------------------------------------------------------- r3887 (orig r2113): jesse | 2005-01-18 10:46:48 -0500 r3885@hualien: jesse | 2005-01-18T15:02:02.574134Z Reimplemented RT::Users->WhoHaveRights to remove O(n^2) SQL calls. (That means it's faster now) Reimplemented SelectOwner widget to take advantage of the new WhoHaveRights ---------------------------------------------------------------------- r3886 (orig r2112): jesse | 2005-01-18 10:46:21 -0500 r3884@hualien: jesse | 2005-01-18T14:59:19.630164Z RT-Ticket: 6352 RT-Status: resolved RT-Update: correspond Fix an invalid custom field acl check ---------------------------------------------------------------------- r3878 (orig r2109): tla | 2005-01-17 20:53:10 -0500 A script to automate testing data creation. Requires a new CPAN module, Text::Lorem. Work in progress. ---------------------------------------------------------------------- r2699 (orig r2072): jesse | 2005-01-13 02:02:25 -0500 Removed a reference to NIS ---------------------------------------------------------------------- r2698 (orig r2071): jesse | 2005-01-12 18:39:03 -0500 RT3.4 schema diagram corrected and updated. ---------------------------------------------------------------------- r2697 (orig r2070): jesse | 2005-01-11 18:58:03 -0500 Readme update to note that manual installation of Apache::DBI may be required. ---------------------------------------------------------------------- r2696 (orig r2069): zev | 2005-01-11 18:49:19 -0500 Fixed spelling and grammar mistakes in etc/RT_Config.pm.in Added some documentation and fixed whitespace in Makefile.in -- From jesse at bestpractical.com Sat Jan 22 15:03:13 2005 From: jesse at bestpractical.com (Jesse Vincent) Date: Sat Jan 22 15:03:18 2005 Subject: [Rt-devel] [PATCH] Quote identifiers. In-Reply-To: <41F2A8F7.9020907@acronis.com> References: <41F2A19E.3050808@acronis.com> <20050122190125.GS24460@bestpractical.com> <41F2A3A1.1000701@acronis.com> <20050122191404.GT24460@bestpractical.com> <41F2A8F7.9020907@acronis.com> Message-ID: <20050122200313.GW24460@bestpractical.com> > Speed. See next benchmark At the expense of clarity. I hate perl's method dispatch performance. Will you update the patch with a comment about _why_ you're doing something that looks like it should be optimized away so that someone else doesn't "fix" it? > #!/usr/bin/perl -w > > > > use strict; > use Benchmark qw(cmpthese); > my $obj = new Dummy; > > > > cmpthese (-6, { > 'with $dbh' => sub { $obj->w_cache(); }, > 'without $dbh' => sub { $obj->wo_cache() }, > }); > > > > package Dummy; > > > > sub new { return bless { dbh => { foo => 'bar' } }, 'Dummy' } > sub dbh { my $self = shift; return $self->{'dbh'} } > > > > sub w_cache { > my $self = shift; > my $dbh = $self->dbh; > for(1..5) { my $x = $dbh->{foo} } > } > sub wo_cache { > my $self = shift; > for(1..5) {my $x = $self->dbh->{foo} } > } > Results: > Rate without $dbh with $dbh > without $dbh 43465/s -- -40% > with $dbh 72127/s 66% -- > > Something similar happens in sub Insert I call quote_identifier in map > context for each column, also dbh sub in SB::Handle more complex then in > benchmark. > > Similar situation in UpdateRecordValue. > -- From rspier at pobox.com Sat Jan 22 15:12:36 2005 From: rspier at pobox.com (Robert Spier) Date: Sat Jan 22 15:25:29 2005 Subject: [Rt-devel] [PATCH] Quote identifiers. In-Reply-To: <41F2A8F7.9020907@acronis.com> References: <41F2A19E.3050808@acronis.com> <20050122190125.GS24460@bestpractical.com> <41F2A3A1.1000701@acronis.com> <20050122191404.GT24460@bestpractical.com> <41F2A8F7.9020907@acronis.com> Message-ID: > Rate without $dbh with $dbh > without $dbh 43465/s -- -40% > with $dbh 72127/s 66% -- > > Something similar happens in sub Insert I call quote_identifier in map > context for each column, also dbh sub in SB::Handle more complex then > in benchmark. > Potential Micro-optimization. How often are these really called? Benchmarks that show improvements in tight loops are great, and I love them, but when the code isn't actually run in a tight loop, it's kind of meaningless. For this kind of statement I'd guess that the time required to actually run the kind of database queries we run greatly dwarfs the time saved by using a temporary. Kind of like a shortcut that saves 5 minutes out of a five hour drive. -R From Ruslan.Zakirov at acronis.com Sat Jan 22 17:54:30 2005 From: Ruslan.Zakirov at acronis.com (Ruslan U. Zakirov) Date: Sat Jan 22 17:58:37 2005 Subject: [Rt-devel] [PATCH] Quote identifiers. In-Reply-To: <20050122200313.GW24460@bestpractical.com> References: <41F2A19E.3050808@acronis.com> <20050122190125.GS24460@bestpractical.com> <41F2A3A1.1000701@acronis.com> <20050122191404.GT24460@bestpractical.com> <41F2A8F7.9020907@acronis.com> <20050122200313.GW24460@bestpractical.com> Message-ID: <41F2D9A6.8070602@acronis.com> Jesse Vincent wrote: > >>Speed. See next benchmark > > > At the expense of clarity. I hate perl's method dispatch > performance. Will you update the patch with a comment about _why_ > you're doing something that looks like it should be optimized away so > that someone else doesn't "fix" it? I resubmit patch where thing looks like you want. Done more complex test that insert 5000 records into SQLite DB. Same time results for both variants. >>#!/usr/bin/perl -w >> >> >> >>use strict; >>use Benchmark qw(cmpthese); >>my $obj = new Dummy; >> >> >> >>cmpthese (-6, { >> 'with $dbh' => sub { $obj->w_cache(); }, >> 'without $dbh' => sub { $obj->wo_cache() }, >> }); >> >> >> >>package Dummy; >> >> >> >>sub new { return bless { dbh => { foo => 'bar' } }, 'Dummy' } >>sub dbh { my $self = shift; return $self->{'dbh'} } >> >> >> >>sub w_cache { >> my $self = shift; >> my $dbh = $self->dbh; >> for(1..5) { my $x = $dbh->{foo} } >>} >>sub wo_cache { >> my $self = shift; >> for(1..5) {my $x = $self->dbh->{foo} } >>} >>Results: >> Rate without $dbh with $dbh >>without $dbh 43465/s -- -40% >>with $dbh 72127/s 66% -- >> >>Something similar happens in sub Insert I call quote_identifier in map >>context for each column, also dbh sub in SB::Handle more complex then in >>benchmark. >> >>Similar situation in UpdateRecordValue. >> > > -------------- next part -------------- A non-text attachment was scrubbed... Name: quote_identifier.patch Type: application/aegis-patch Size: 1245 bytes Desc: not available Url : http://bestpractical.com/pipermail/rt-devel/attachments/20050123/6d47fbc8/quote_identifier-0001.bin From Ruslan.Zakirov at acronis.com Sun Jan 23 14:56:42 2005 From: Ruslan.Zakirov at acronis.com (Ruslan U. Zakirov) Date: Sun Jan 23 15:00:52 2005 Subject: [Rt-devel] [PATCH] DBIx::SB tests generalization Message-ID: <41F4017A.80506@acronis.com> Hello. Attached patch is ready to apply, but part of the work to make SB tests more generic. t/01basic.t - introduces some basic tests, like different handle loading and testing that constructor return right objects. Now you don't need to duplicate that tests in other test files. t/utils.pl - is library of utils required for tests. Now it has: @SupportedDrivers - list of supported back ends get_handle - function that returns new handle by name Regards. Ruslan. -------------- next part -------------- A non-text attachment was scrubbed... Name: generalize_tests_round1.patch Type: application/aegis-patch Size: 4296 bytes Desc: not available Url : http://bestpractical.com/pipermail/rt-devel/attachments/20050123/6969af30/generalize_tests_round1.bin From wash at wananchi.com Mon Jan 24 04:00:03 2005 From: wash at wananchi.com (Odhiambo Washington) Date: Mon Jan 24 04:17:24 2005 Subject: [Rt-devel] How to fix RTIR + rt-3.4.0rc6? In-Reply-To: <20050122112618.GF51708@ns2.wananchi.com> References: <20050122112618.GF51708@ns2.wananchi.com> Message-ID: <20050124090003.GF25493@ns2.wananchi.com> Hello experts, I have made two attempts at running RTIR 1.1.3 The test box is FreeBSD 5.3-RELEASE, Perl-5.8.5, Apache-2.0.25, mod-perl-2 (yes!) rt-3.0.4rc6 and MySQL 4.1. The other box is FreeBSD 4.11-STABLE, Perl-5.8.5, Apache-1.3.33, mod-perl-1.29, rt-3.2.22 and MySQL 4.0. In both cases, I end up with the same error, but I have no idea how to fix it. I'd appreciate any help. error: Error during compilation of /opt/rt3/share/html/Callbacks/RTIR/autohandler/Default: Can't locate RT/IR.pm in @INC (@INC contains: /opt/rt3/local/lib /opt/rt3/lib /usr/local/lib/perl5/site_perl/5.8.5/mach/Apache2 /usr/local/lib/perl5/site_perl/5.8.5/mach /usr/local/lib/perl5/site_perl/5.8.5 /usr/local/lib/perl5/site_perl /usr/local/lib/perl5/5.8.5/BSDPAN /usr/local/lib/perl5/5.8.5/mach /usr/local/lib/perl5/5.8.5 /usr/local/ /usr/local/lib/perl) at /opt/rt3/share/html/Callbacks/RTIR/autohandler/Default line 2. Stack: [/opt/rt3/share/html/Callbacks/RTIR/autohandler/Default:2] [RT/IR.pm:2] [/usr/local/lib/perl5/site_perl/5.8.5/HTML/Mason/Interp.pm:543] [/usr/local/lib/perl5/site_perl/5.8.5/HTML/Mason/Interp.pm:335] [/usr/local/lib/perl5/site_perl/5.8.5/HTML/Mason/Request.pm:892] [/usr/local/lib/perl5/site_perl/5.8.5/HTML/Mason/Request.pm:996] [/opt/rt3/share/html/Elements/Callback:70] [/opt/rt3/share/html/autohandler:206] BEGIN failed--compilation aborted at /opt/rt3/share/html/Callbacks/RTIR/autohandler/Default line 2. Stack: [/opt/rt3/share/html/Callbacks/RTIR/autohandler/Default:2] [/usr/local/lib/perl5/site_perl/5.8.5/HTML/Mason/Interp.pm:543] [/usr/local/lib/perl5/site_perl/5.8.5/HTML/Mason/Interp.pm:335] [/usr/local/lib/perl5/site_perl/5.8.5/HTML/Mason/Request.pm:892] [/usr/local/lib/perl5/site_perl/5.8.5/HTML/Mason/Request.pm:996] [/opt/rt3/share/html/Elements/Callback:70] [/opt/rt3/share/html/autohandler:206] context: 1: <%init> 2: use RT::IR; 3: return(1); 4: 5: code stack: /usr/local/lib/perl5/site_perl/5.8.5/HTML/Mason/Interp.pm:336 /usr/local/lib/perl5/site_perl/5.8.5/HTML/Mason/Request.pm:892 /usr/local/lib/perl5/site_perl/5.8.5/HTML/Mason/Request.pm:996 /opt/rt3/share/html/Elements/Callback:70 /opt/rt3/share/html/autohandler:206 cheers - wash +----------------------------------+-----------------------------------------+ Odhiambo Washington . WANANCHI ONLINE LTD (Nairobi, KE) | wash _at_ wananchi _ dot _ com . 1ere Etage, Loita Hse, Loita St., | GSM: (+254) 722 743 223 . # 10286, 00100 NAIROBI | GSM: (+254) 733 744 121 . (+254) 020 313 985 - 9 | +---------------------------------+------------------------------------------+ "Oh My God! They killed init! You Bastards!" --from a /. post From jesse at bestpractical.com Mon Jan 24 07:22:08 2005 From: jesse at bestpractical.com (Jesse Vincent) Date: Mon Jan 24 07:22:15 2005 Subject: [Rt-devel] How to fix RTIR + rt-3.4.0rc6? In-Reply-To: <20050124090003.GF25493@ns2.wananchi.com> References: <20050122112618.GF51708@ns2.wananchi.com> <20050124090003.GF25493@ns2.wananchi.com> Message-ID: <20050124122208.GB24460@bestpractical.com> On Mon, Jan 24, 2005 at 12:00:03PM +0300, Odhiambo Washington wrote: > > Hello experts, > > I have made two attempts at running RTIR 1.1.3 Give 1.1.4 a shot? > The test box is FreeBSD 5.3-RELEASE, Perl-5.8.5, Apache-2.0.25, > mod-perl-2 (yes!) rt-3.0.4rc6 and MySQL 4.1. > > The other box is FreeBSD 4.11-STABLE, Perl-5.8.5, Apache-1.3.33, > mod-perl-1.29, rt-3.2.22 and MySQL 4.0. > > In both cases, I end up with the same error, but I have no idea > how to fix it. > I'd appreciate any help. > > > > error: Error during compilation of /opt/rt3/share/html/Callbacks/RTIR/autohandler/Default: > Can't locate RT/IR.pm in @INC (@INC contains: /opt/rt3/local/lib /opt/rt3/lib /usr/local/lib/perl5/site_perl/5.8.5/mach/Apache2 /usr/local/lib/perl5/site_perl/5.8.5/mach /usr/local/lib/perl5/site_perl/5.8.5 /usr/local/lib/perl5/site_perl /usr/local/lib/perl5/5.8.5/BSDPAN /usr/local/lib/perl5/5.8.5/mach /usr/local/lib/perl5/5.8.5 /usr/local/ /usr/local/lib/perl) at /opt/rt3/share/html/Callbacks/RTIR/autohandler/Default line 2. > > Stack: > [/opt/rt3/share/html/Callbacks/RTIR/autohandler/Default:2] > [RT/IR.pm:2] > [/usr/local/lib/perl5/site_perl/5.8.5/HTML/Mason/Interp.pm:543] > [/usr/local/lib/perl5/site_perl/5.8.5/HTML/Mason/Interp.pm:335] > [/usr/local/lib/perl5/site_perl/5.8.5/HTML/Mason/Request.pm:892] > [/usr/local/lib/perl5/site_perl/5.8.5/HTML/Mason/Request.pm:996] > [/opt/rt3/share/html/Elements/Callback:70] > [/opt/rt3/share/html/autohandler:206] > BEGIN failed--compilation aborted at /opt/rt3/share/html/Callbacks/RTIR/autohandler/Default line 2. > > Stack: > [/opt/rt3/share/html/Callbacks/RTIR/autohandler/Default:2] > [/usr/local/lib/perl5/site_perl/5.8.5/HTML/Mason/Interp.pm:543] > [/usr/local/lib/perl5/site_perl/5.8.5/HTML/Mason/Interp.pm:335] > [/usr/local/lib/perl5/site_perl/5.8.5/HTML/Mason/Request.pm:892] > [/usr/local/lib/perl5/site_perl/5.8.5/HTML/Mason/Request.pm:996] > [/opt/rt3/share/html/Elements/Callback:70] > [/opt/rt3/share/html/autohandler:206] > > context: > 1: <%init> > 2: use RT::IR; > 3: return(1); > 4: > 5: > code stack: /usr/local/lib/perl5/site_perl/5.8.5/HTML/Mason/Interp.pm:336 > /usr/local/lib/perl5/site_perl/5.8.5/HTML/Mason/Request.pm:892 > /usr/local/lib/perl5/site_perl/5.8.5/HTML/Mason/Request.pm:996 > /opt/rt3/share/html/Elements/Callback:70 > /opt/rt3/share/html/autohandler:206 > > > > > cheers > - wash > +----------------------------------+-----------------------------------------+ > Odhiambo Washington . WANANCHI ONLINE LTD (Nairobi, KE) | > wash _at_ wananchi _ dot _ com . 1ere Etage, Loita Hse, Loita St., | > GSM: (+254) 722 743 223 . # 10286, 00100 NAIROBI | > GSM: (+254) 733 744 121 . (+254) 020 313 985 - 9 | > +---------------------------------+------------------------------------------+ > "Oh My God! They killed init! You Bastards!" > --from a /. post > _______________________________________________ > Rt-devel mailing list > Rt-devel@lists.bestpractical.com > http://lists.bestpractical.com/cgi-bin/mailman/listinfo/rt-devel > -- From lucas at dextra.com.br Mon Jan 24 08:22:12 2005 From: lucas at dextra.com.br (Lucas Fernando Rosada) Date: Mon Jan 24 08:26:26 2005 Subject: [Rt-devel] Configuration Extraction Tool Message-ID: <20050124132212.17704.qmail@hm28.locaweb.com.br> Hello, does anyone know any RT tool that generate an initialdata file from the data that's into the database? Thaks, Lucas F. Rosada lucas@dextra.com.br From jesse at bestpractical.com Mon Jan 24 15:31:24 2005 From: jesse at bestpractical.com (Jesse Vincent) Date: Mon Jan 24 15:31:31 2005 Subject: [Rt-devel] Configuration Extraction Tool In-Reply-To: <20050124132212.17704.qmail@hm28.locaweb.com.br> References: <20050124132212.17704.qmail@hm28.locaweb.com.br> Message-ID: <20050124203124.GD24460@bestpractical.com> On Mon, Jan 24, 2005 at 11:22:12AM -0200, Lucas Fernando Rosada wrote: > > Hello, does anyone know any RT tool that generate an initialdata file from the data that's into the database? > rt-dump-database in RT 3.4 Jesse > Thaks, > > Lucas F. Rosada > lucas@dextra.com.br > _______________________________________________ > Rt-devel mailing list > Rt-devel@lists.bestpractical.com > http://lists.bestpractical.com/cgi-bin/mailman/listinfo/rt-devel > -- From javila at netlojix.com Mon Jan 24 17:52:21 2005 From: javila at netlojix.com (javila@netlojix.com) Date: Mon Jan 24 18:02:24 2005 Subject: [Rt-devel] Overlay Issues Message-ID: <1106607141.41f57c255ffdf@webmail.netlojix.com> After many many attempts at trying to get things to work, i just cant seem to figure out how get things going (Maybe i need to up the ammount of daily tequilla intake)... So i have rt3 installed in /opt/rt3/ I try to Create an overlay for queue.pm... so i copy lib/RT/Queue.pm to local/lib/RT/Queue_Local.pm and i strip out all but the necessities. === BOF === use strict; no warnings qw(redefine); sub Create { my $self = shift; my %args = ( ... return ( 0, $self->loc("WTF") ); } 1; === EOF === I bounce apache all is happy... but the code change does not seem to take effect. If there is a syntax error in this file apache will not bounce so obviously it knows about its existance. Thank you for the help. From rt at chaka.net Mon Jan 24 18:16:23 2005 From: rt at chaka.net (Todd Chapman) Date: Mon Jan 24 18:59:20 2005 Subject: [Rt-devel] Overlay Issues In-Reply-To: <1106607141.41f57c255ffdf@webmail.netlojix.com> References: <1106607141.41f57c255ffdf@webmail.netlojix.com> Message-ID: <20050124231623.GM27116@chaka.net> What is $LocalPath set to in RT.pm? On Mon, Jan 24, 2005 at 02:52:21PM -0800, javila@netlojix.com wrote: > After many many attempts at trying to get things to work, i just cant seem to > figure out how get things going (Maybe i need to up the ammount of daily > tequilla intake)... So i have rt3 installed in /opt/rt3/ I try to Create an > overlay for queue.pm... so i copy lib/RT/Queue.pm to local/lib/RT/Queue_Local.pm > and i strip out all but the necessities. > > === BOF === > use strict; > no warnings qw(redefine); > sub Create { > my $self = shift; > my %args = ( > ... > return ( 0, $self->loc("WTF") ); > } > 1; > === EOF === > I bounce apache all is happy... but the code change does not seem to take > effect. If there is a syntax error in this file apache will not bounce so > obviously it knows about its existance. Thank you for the help. > _______________________________________________ > Rt-devel mailing list > Rt-devel@lists.bestpractical.com > http://lists.bestpractical.com/cgi-bin/mailman/listinfo/rt-devel From lucas at dextra.com.br Tue Jan 25 06:04:50 2005 From: lucas at dextra.com.br (Lucas Fernando Rosada) Date: Tue Jan 25 06:09:25 2005 Subject: [Rt-devel] Configuration Extraction Tool In-Reply-To: <20050124203124.GD24460@bestpractical.com> References: <20050124203124.GD24460@bestpractical.com> Message-ID: <20050125110450.550.qmail@hm28.locaweb.com.br> Thanks, Jesse. I have another doubt: does the factory script (rt-path/sbin/factory) supports Postgres database? When I tried to execute it with a PG database, it crashed when the DESCRIBE commanda was tried (my $sth = $dbh->prepare("DESCRIBE $table")). Thanks a lot, Lucas F. Rosada lucas@dextra.com.br On Mon, 24 Jan 2005 15:31:24 -0500, Jesse Vincent escreveu: > De: Jesse Vincent > Data: Mon, 24 Jan 2005 15:31:24 -0500 > Para: Lucas Fernando Rosada > Assunto: Re: [Rt-devel] Configuration Extraction Tool > > > On Mon, Jan 24, 2005 at 11:22:12AM -0200, Lucas Fernando Rosada wrote: > > > > Hello, does anyone know any RT tool that generate an initialdata file from the data that's into the database? > > > > rt-dump-database in RT 3.4 > > Jesse > > > Thaks, > > > > Lucas F. Rosada > > lucas@dextra.com.br > > _______________________________________________ > > Rt-devel mailing list > > Rt-devel@lists.bestpractical.com > > http://lists.bestpractical.com/cgi-bin/mailman/listinfo/rt-devel > > > > -- > > > From lucas at dextra.com.br Tue Jan 25 06:29:37 2005 From: lucas at dextra.com.br (Lucas Fernando Rosada) Date: Tue Jan 25 06:33:58 2005 Subject: [Rt-devel] Re: Oberlay Issues Message-ID: <20050125112937.31331.qmail@hm28.locaweb.com.br> Hi, guy! I had the same problem and I discovered that the overlays scope (for the *_Vendor and *_Local files) are in the current directory. If you put an loca/lib/RT/Queue_Local.pm, you're going to "overwrite" the lib/RT/Queue_Local.pm: you're not doing the overlay for Queue.pm. So, you need to copy the lib/RT/Queue.pm to local/lib/RT/Queue.pm, and only after the local copy you develop and see the effect of your overlay. If you want, you could remove all the code from the local/lib/Queue.pm, but it's necessary to keep the final lines of the file, that activate the File_Vender or File_Local if it exists. regards Lucas F. ROsada lucas@dextra.com.br From lucas at dextra.com.br Tue Jan 25 07:44:55 2005 From: lucas at dextra.com.br (Lucas Fernando Rosada) Date: Tue Jan 25 07:49:17 2005 Subject: [Rt-devel] Re: Oberlay Issues In-Reply-To: References: Message-ID: <20050125124455.4782.qmail@hm28.locaweb.com.br> I'm sorry, I just discovered you need to keep a copy of lib/RT/Queue.pm without touching in the loca/lib/RT dir. You MUST NOT touch the file, cause it will "overwrite" the original in lib/RT/. Right? Thanks, Lucas F. Rosada lucas@dextra.com.br On Tue, 25 Jan 2005 09:29:37 -0200, "Lucas Fernando Rosada" escreveu: > De: "Lucas Fernando Rosada" > Data: Tue, 25 Jan 2005 09:29:37 -0200 > Para: javila@netlojix.com > Assunto: Re: Oberlay Issues > > > > Hi, guy! > I had the same problem and I discovered that the overlays scope (for the *_Vendor and *_Local files) are in the current directory. If you put an loca/lib/RT/Queue_Local.pm, you're going to "overwrite" the lib/RT/Queue_Local.pm: you're not doing the overlay for Queue.pm. > So, you need to copy the lib/RT/Queue.pm to local/lib/RT/Queue.pm, and only after the local copy you develop and see the effect of your overlay. If you want, you could remove all the code from the local/lib/Queue.pm, but it's necessary to keep the final lines of the file, that activate the File_Vender or File_Local if it exists. > > regards > > Lucas F. ROsada > lucas@dextra.com.br > > > From rt at chaka.net Tue Jan 25 16:10:26 2005 From: rt at chaka.net (Todd Chapman) Date: Tue Jan 25 16:53:19 2005 Subject: [Rt-devel] AssetTracker 0.0.1 Release early, release often! Message-ID: <20050125211026.GX27116@chaka.net> My earlier post was toi big (100KB). If anyone wants the tar file let me know where to mail it to. -Todd -------------------------- Here is my initial shot at AssetTracker, an asset tracking extension to RT 3.4. Don't be suprised if it doesn't work for you. :) I am sending this out to get some feedback on design. There are many bugs. Click on some menu items and you will get an error becuase those components don't exist yet. Hopefully this will generate some interest. See the README in the tar file. -Todd From chris at meta4-group.com Wed Jan 26 01:29:42 2005 From: chris at meta4-group.com (Chris Li) Date: Wed Jan 26 01:32:50 2005 Subject: [Rt-devel] Native Japanese Support Message-ID: <009001c50370$6c0da060$9900a8c0@hfwsrv1> Hi, We have been using the system for quite a long time for English and Chinese, but when it comes to the time to use it for Japanese, it simply crap out. It seems that RT cannot detect the incoming message as accurate as Chinese and the email it sends cannot be cheched by many Japanese based web client. Anyone having the same experience with me and able to share a solution with me? I am in the nightmare of it. Regards, Chris Li System Administartor Meta4 Group Limited -------------- next part -------------- An HTML attachment was scrubbed... URL: http://bestpractical.com/pipermail/rt-devel/attachments/20050126/7ec0f396/attachment.htm From jesse at bestpractical.com Wed Jan 26 03:42:03 2005 From: jesse at bestpractical.com (Jesse Vincent) Date: Wed Jan 26 03:42:08 2005 Subject: [Rt-devel] Native Japanese Support In-Reply-To: <009001c50370$6c0da060$9900a8c0@hfwsrv1> References: <009001c50370$6c0da060$9900a8c0@hfwsrv1> Message-ID: <20050126084202.GE24460@bestpractical.com> > > We have been using the system for quite a long time for English and Chinese, but when it comes to the time to use it for Japanese, it simply crap out. It seems that RT cannot detect the incoming message as accurate as Chinese and the email it sends cannot be cheched by many Japanese based web client. > > Anyone having the same experience with me and able to share a solution with me? I am in the nightmare of it. > Can you tell us a bit more about which versions of RT, perl, apache, etc. you're running and what "craps out" means in your case? Jesse > Regards, > > Chris Li > System Administartor > Meta4 Group Limited > _______________________________________________ > Rt-devel mailing list > Rt-devel@lists.bestpractical.com > http://lists.bestpractical.com/cgi-bin/mailman/listinfo/rt-devel -- From rt at chaka.net Wed Jan 26 13:51:34 2005 From: rt at chaka.net (Todd Chapman) Date: Wed Jan 26 14:34:35 2005 Subject: [Rt-devel] Announce: AssetTracker wiki and mailing lists. Message-ID: <20050126185134.GG27116@chaka.net> I'm not sure I am ready to make a general announcement yet, but being on rt-devel comes with privileges. :) The AT wiki is at: http://atwiki.chaka.net/ Warning, the current server is VERY slow. I think maybe I should pick up a Mac mini to handle the load. :) The AT mailing lists are: at-announce@lists.chaka.net at-devel@lists.chaka.net at-users@lists.chaka.net Also AT now can be downloaded from the wiki, and from CPAN in the no-too-distant future. -Todd From wash at wananchi.com Wed Jan 26 11:55:21 2005 From: wash at wananchi.com (Odhiambo Washington) Date: Wed Jan 26 17:43:49 2005 Subject: [Rt-devel] How to "clean" a DB of RTIR data Message-ID: <20050126165521.GA21404@ns2.wananchi.com> I need to know if it is possible to completely cleanup a database of any rtir data. Why is this? In the process of trying out rtir, I tested on two machines. On the test machine, I have no problem. I can wipe out the database and recreate everything from scratch. My tests on the "staging" box failed, and in an attempt to reproduce the results, I also tested on a production box, after I made a backup of the database and RT_BASE. However the production system DB continued to receive RT mail, so if I try to restore the database, I will loose a couple tickets in RT. That is simply unacceptable. That is why I would like to know if I could clean the DB of all RTIR data, then do a completely "new" RTIR installation, not an upgrade. I have tested rtir-1.1.4 on my staging box and it did install cleanly, as opposed to the previous 1.1.3 version. Is it possible to do this? It is possible to "reverse" what `make initdb` does during rtir installation? cheers - wash +----------------------------------+-----------------------------------------+ Odhiambo Washington . WANANCHI ONLINE LTD (Nairobi, KE) | wash _at_ wananchi _ dot _ com . 1ere Etage, Loita Hse, Loita St., | GSM: (+254) 722 743 223 . # 10286, 00100 NAIROBI | GSM: (+254) 733 744 121 . (+254) 020 313 985 - 9 | +---------------------------------+------------------------------------------+ "Oh My God! They killed init! You Bastards!" --from a /. post From jralbert at uvic.ca Wed Jan 26 19:43:36 2005 From: jralbert at uvic.ca (Jeff Albert) Date: Wed Jan 26 20:01:08 2005 Subject: [Rt-devel] RT 3.4 on OpenBSD Message-ID: <200501270046.j0R0jeI82343164@castle.comp.uvic.ca> Hi all, Just curious: are people running RT on OpenBSD? There are no specific install docs for it (that I could find) so I'm sort of winging it based on the general documentation. Are there any specific tips you fine folks might bring forth before I rush headlong into installing? Cheers, Jeff Albert Departmental Computing Support, University of Victoria jralbert@uvic.ca -------------- next part -------------- An HTML attachment was scrubbed... URL: http://bestpractical.com/pipermail/rt-devel/attachments/20050126/0f5ba3c6/attachment.htm From steve at fisharerojo.org Wed Jan 26 21:02:11 2005 From: steve at fisharerojo.org (Steve Peters) Date: Wed Jan 26 21:17:52 2005 Subject: [Rt-devel] RT 3.4 on OpenBSD In-Reply-To: <200501270046.j0R0jeI82343164@castle.comp.uvic.ca> References: <200501270046.j0R0jeI82343164@castle.comp.uvic.ca> Message-ID: <20050127020211.GA11143@mccoy.peters.homeunix.org> On Wed, Jan 26, 2005 at 04:43:36PM -0800, Jeff Albert wrote: > Hi all, > > Just curious: are people running RT on OpenBSD? There are no specific > install docs for it (that I could find) so I'm sort of winging it based on > the general documentation. Are there any specific tips you fine folks might > bring forth before I rush headlong into installing? > > > > Cheers, > > Jeff Albert > > Departmental Computing Support, > > University of Victoria > > jralbert@uvic.ca > > _______________________________________________ > Rt-devel mailing list > Rt-devel@lists.bestpractical.com > http://lists.bestpractical.com/cgi-bin/mailman/listinfo/rt-devel I'm running a release candidate right now on OpenBSD without any problems. There a few things to watch for. First, the default Apache on OpenBSD runs in a chroot'ed environment, but if you are using mod_perl and have Apache configured correctly, its not a problem. The only other issue I had was with the connection to syslog, but I think that was more of an issue between the keyboard and chair than anything with OpenBSD. Just to be sure, reboot after you OpenBSD install (if this is a new server) and before you install RT. Steve Peters steve@fisharerojo.org From chris at meta4-group.com Thu Jan 27 02:53:23 2005 From: chris at meta4-group.com (Chris Li) Date: Thu Jan 27 02:56:40 2005 Subject: Fw: [Rt-devel] Native Japanese Support Message-ID: <01a701c50445$47318790$9900a8c0@hfwsrv1> ----- Original Message ----- From: "Chris Li" To: "Jesse Vincent" Sent: Wednesday, January 26, 2005 4:53 PM Subject: Re: [Rt-devel] Native Japanese Support > Hi! > ----- Original Message ----- > From: "Jesse Vincent" > To: "Chris Li" > Cc: > Sent: Wednesday, January 26, 2005 4:42 PM > Subject: Re: [Rt-devel] Native Japanese Support > > > > > > > > > > We have been using the system for quite a long time for English and > Chinese, but when it comes to the time to use it for Japanese, it simply > crap out. It seems that RT cannot detect the incoming message as accurate as > Chinese and the email it sends cannot be cheched by many Japanese based web > client. > > > > > > Anyone having the same experience with me and able to share a > solution with me? I am in the nightmare of it. > > > > > > > Can you tell us a bit more about which versions of RT, perl, apache, > > etc. you're running and > > what "craps out" means in your case? > > > We are using RT 3.2.2, Perl v5.8.3 and apache 2.0.51 on a fedora core 2 box > The system response Japanese with messed code (my gut feeling is that it > cannot detect what is the origional code) > > Chris > > > > Jesse > > > > > > > Regards, > > > > > > Chris Li > > > System Administartor > > > Meta4 Group Limited > > > > > _______________________________________________ > > > Rt-devel mailing list > > > Rt-devel@lists.bestpractical.com > > > http://lists.bestpractical.com/cgi-bin/mailman/listinfo/rt-devel > > > > > > -- > From JRamon.Alfageme at alhambra-eidos.com Thu Jan 27 06:10:28 2005 From: JRamon.Alfageme at alhambra-eidos.com (=?iso-8859-1?Q?Juan_Ram=F3n_Alfageme_Mata?=) Date: Thu Jan 27 06:27:03 2005 Subject: [Rt-devel] (no subject) Message-ID: <1515E176F4AD2E4AA41D92D7C006BC47015A9C38@ALHM4SRV.alhpridom.as> -------------- next part -------------- An HTML attachment was scrubbed... URL: http://bestpractical.com/pipermail/rt-devel/attachments/20050127/8e48fea2/attachment.html From JRamon.Alfageme at alhambra-eidos.com Thu Jan 27 07:26:45 2005 From: JRamon.Alfageme at alhambra-eidos.com (=?iso-8859-1?Q?Juan_Ram=F3n_Alfageme_Mata?=) Date: Thu Jan 27 07:31:20 2005 Subject: [Rt-devel] RTIR_Config.pm error starting apache Message-ID: <1515E176F4AD2E4AA41D92D7C006BC47015A9C59@ALHM4SRV.alhpridom.as> I have a problem every time I start apache, I get the following trace: Starting web server: apacheUse of uninitialized value in concatenation (.) or string at /opt/rt3/local/etc/IR/RTIR_Config.pm line 3. Use of uninitialized value in concatenation (.) or string at /opt/rt3/local/etc/IR/RTIR_Config.pm line 115. Use of uninitialized value in concatenation (.) or string at /opt/rt3/local/etc/IR/RTIR_Config.pm line 115. Use of uninitialized value in concatenation (.) or string at /opt/rt3/local/etc/IR/RTIR_Config.pm line 115. Use of uninitialized value in concatenation (.) or string at /opt/rt3/local/etc/IR/RTIR_Config.pm line 115. Use of uninitialized value in concatenation (.) or string at /opt/rt3/local/etc/IR/RTIR_Config.pm line 115. Use of uninitialized value in concatenation (.) or string at /opt/rt3/local/etc/IR/RTIR_Config.pm line 115. Use of uninitialized value in concatenation (.) or string at /opt/rt3/local/etc/IR/RTIR_Config.pm line 115. Use of uninitialized value in concatenation (.) or string at /opt/rt3/local/etc/IR/RTIR_Config.pm line 115. Use of uninitialized value in concatenation (.) or string at /opt/rt3/local/etc/IR/RTIR_Config.pm line 115. Use of uninitialized value in concatenation (.) or string at /opt/rt3/local/etc/IR/RTIR_Config.pm line 115. Use of uninitialized value in concatenation (.) or string at /opt/rt3/local/etc/IR/RTIR_Config.pm line 115. Use of uninitialized value in concatenation (.) or string at /opt/rt3/local/etc/IR/RTIR_Config.pm line 115. Use of uninitialized value in concatenation (.) or string at /opt/rt3/local/etc/IR/RTIR_Config.pm line 115. Use of uninitialized value in concatenation (.) or string at /opt/rt3/local/etc/IR/RTIR_Config.pm line 115. Use of uninitialized value in concatenation (.) or string at /opt/rt3/local/etc/IR/RTIR_Config.pm line 115. Use of uninitialized value in concatenation (.) or string at /opt/rt3/local/etc/IR/RTIR_Config.pm line 115. Use of uninitialized value in concatenation (.) or string at /opt/rt3/local/etc/IR/RTIR_Config.pm line 115. Use of uninitialized value in concatenation (.) or string at /opt/rt3/local/etc/IR/RTIR_Config.pm line 115. Use of uninitialized value in concatenation (.) or string at /opt/rt3/local/etc/IR/RTIR_Config.pm line 115. Use of uninitialized value in concatenation (.) or string at /opt/rt3/local/etc/IR/RTIR_Config.pm line 115. Use of uninitialized value in concatenation (.) or string at /opt/rt3/local/etc/IR/RTIR_Config.pm line 115. Use of uninitialized value in concatenation (.) or string at /opt/rt3/local/etc/IR/RTIR_Config.pm line 115. Use of uninitialized value in concatenation (.) or string at /opt/rt3/local/etc/IR/RTIR_Config.pm line 115. Use of uninitialized value in concatenation (.) or string at /opt/rt3/local/etc/IR/RTIR_Config.pm line 115. Use of uninitialized value in concatenation (.) or string at /opt/rt3/local/etc/IR/RTIR_Config.pm line 115. Use of uninitialized value in concatenation (.) or string at /opt/rt3/local/etc/IR/RTIR_Config.pm line 115. Use of uninitialized value in concatenation (.) or string at /opt/rt3/local/etc/IR/RTIR_Config.pm line 115. Use of uninitialized value in concatenation (.) or string at /opt/rt3/local/etc/IR/RTIR_Config.pm line 115. That first error was fixed changing "," with "." Set($rtirname , "RTIR for " . $rtname); --> Set($rtirname , "RTIR for " , $rtname); But I still get the line 115 error (Default formats for RTIR search results), RT works fine but in RTIR menu if you select any queue you don't get any result nothing is displayed, apache start trace follows: Starting web server: apacheUse of uninitialized value in concatenation (.) or string at /opt/rt3/local/etc/IR/RTIR_Config.pm line 115. Use of uninitialized value in concatenation (.) or string at /opt/rt3/local/etc/IR/RTIR_Config.pm line 115. Use of uninitialized value in concatenation (.) or string at /opt/rt3/local/etc/IR/RTIR_Config.pm line 115. Use of uninitialized value in concatenation (.) or string at /opt/rt3/local/etc/IR/RTIR_Config.pm line 115. Use of uninitialized value in concatenation (.) or string at /opt/rt3/local/etc/IR/RTIR_Config.pm line 115. Use of uninitialized value in concatenation (.) or string at /opt/rt3/local/etc/IR/RTIR_Config.pm line 115. Use of uninitialized value in concatenation (.) or string at /opt/rt3/local/etc/IR/RTIR_Config.pm line 115. Use of uninitialized value in concatenation (.) or string at /opt/rt3/local/etc/IR/RTIR_Config.pm line 115. Use of uninitialized value in concatenation (.) or string at /opt/rt3/local/etc/IR/RTIR_Config.pm line 115. Use of uninitialized value in concatenation (.) or string at /opt/rt3/local/etc/IR/RTIR_Config.pm line 115. Use of uninitialized value in concatenation (.) or string at /opt/rt3/local/etc/IR/RTIR_Config.pm line 115. Use of uninitialized value in concatenation (.) or string at /opt/rt3/local/etc/IR/RTIR_Config.pm line 115. Use of uninitialized value in concatenation (.) or string at /opt/rt3/local/etc/IR/RTIR_Config.pm line 115. Use of uninitialized value in concatenation (.) or string at /opt/rt3/local/etc/IR/RTIR_Config.pm line 115. Use of uninitialized value in concatenation (.) or string at /opt/rt3/local/etc/IR/RTIR_Config.pm line 115. Use of uninitialized value in concatenation (.) or string at /opt/rt3/local/etc/IR/RTIR_Config.pm line 115. Use of uninitialized value in concatenation (.) or string at /opt/rt3/local/etc/IR/RTIR_Config.pm line 115. Use of uninitialized value in concatenation (.) or string at /opt/rt3/local/etc/IR/RTIR_Config.pm line 115. Use of uninitialized value in concatenation (.) or string at /opt/rt3/local/etc/IR/RTIR_Config.pm line 115. Use of uninitialized value in concatenation (.) or string at /opt/rt3/local/etc/IR/RTIR_Config.pm line 115. Use of uninitialized value in concatenation (.) or string at /opt/rt3/local/etc/IR/RTIR_Config.pm line 115. Use of uninitialized value in concatenation (.) or string at /opt/rt3/local/etc/IR/RTIR_Config.pm line 115. Use of uninitialized value in concatenation (.) or string at /opt/rt3/local/etc/IR/RTIR_Config.pm line 115. Use of uninitialized value in concatenation (.) or string at /opt/rt3/local/etc/IR/RTIR_Config.pm line 115. Use of uninitialized value in concatenation (.) or string at /opt/rt3/local/etc/IR/RTIR_Config.pm line 115. Use of uninitialized value in concatenation (.) or string at /opt/rt3/local/etc/IR/RTIR_Config.pm line 115. Use of uninitialized value in concatenation (.) or string at /opt/rt3/local/etc/IR/RTIR_Config.pm line 115. Use of uninitialized value in concatenation (.) or string at /opt/rt3/local/etc/IR/RTIR_Config.pm line 115. I am using RT 3.4rc6 with RTIR 1.1.5, firstly I tried 1.1.4 but when I get the line 115 error I tried next version (1.1.5) with the same result. Any help would be greatly appreciated. From stefan at van-boxmer.de Thu Jan 27 12:04:11 2005 From: stefan at van-boxmer.de (Stefan Fischer) Date: Thu Jan 27 12:29:11 2005 Subject: [Rt-devel] RT 3.2.2 custom scrip with user defined condition fails In-Reply-To: <1515E176F4AD2E4AA41D92D7C006BC47015A9C59@ALHM4SRV.alhpridom.as> References: <1515E176F4AD2E4AA41D92D7C006BC47015A9C59@ALHM4SRV.alhpridom.as> Message-ID: <41F91F0B.5000800@van-boxmer.de> Hello, i have asked this problem a view days bevore in the rt-users section, but nobody could help me. I have a custom condition if (($self->TransactionObj->Type eq "CustomField") and ($self->TransactionObj->Field eq 2) and ($self->TransactionObj->NewValue eq "20 Antrag eingetroffen")) { return(1); } else { return(undef); } in a user defined (global) scrip to send out email (global template) on custom field change. But even if i try only "return(1);" as condition, the error is the same. No email is send and in syslog the followinf error apears: Jan 27 17:48:39 debian64 RT: Scrip Prepare 92 died. - $/ as given is currently unsupported at /usr/share/perl5/MIME/Body.pm line 194 Stack: [/usr/share/perl/5.8/Carp.pm:86] [/usr/share/perl5/IO/ScalarArray.pm:350] [/usr/share/perl5/MIME/Body.pm:194] [/usr/share/request-tracker3.2/lib/RT/I18N.pm:218] [/usr/share/request-tracker3.2/lib/RT/Action/SendEmail.pm:153] [/usr/share/request-tracker3.2/lib/RT/Action/Notify.pm:65] [/usr/share/request-tracker3.2/lib/RT/ScripAction_Overlay.pm:217] [/usr/share/request-tracker3.2/lib/RT/Scrip_Overlay.pm:468] [/usr/share/request-tracker3.2/lib/RT/Scrips_Overlay.pm:232] [/usr/share/request-tracker3.2/lib/RT/Transaction_Overlay.pm:146] [/usr/share/request-tracker3.2/lib/RT/Ticket_Overlay.pm:3707] [/usr/share/request-tracker3.2/lib/RT/Ticket_Overlay.pm:3471] [/usr/share/request-tracker3.2/lib/RT/Ticket_Overlay.pm:3379] [/usr/share/request-tracker3.2/lib/RT/Interface/Web.pm:1153] [/usr/share/request-tracker3.2/html/Ticket/Modify.html:71] [/usr/share/request-tracker3 I am totally lost. Hopefully somebody can help me. Greetings, Stefan From matthew.watson at staff.netspace.net.au Thu Jan 27 20:45:06 2005 From: matthew.watson at staff.netspace.net.au (Matthew Watson) Date: Thu Jan 27 20:50:47 2005 Subject: [Rt-devel] Per user UI customisation Message-ID: <56F211C5E3F24F47B103EA1B253822BE460F7F@vic-cr-ex1.staff.netspace.net.au> Morning.. I'm just wondering what is happening with per user UI customisation . Is anyone working on this? The RT API seems to have everything in it to enabled this (eg, Attributes), so, before I start hacking around too much, is this something that is in the works? Regards -- Matthew Watson Netspace Online Systems (03) 98110010 This email and any files transmitted with it are confidential and intended solely for the use of the individual or entity to whom they are addressed. Please notify the sender immediately by email if you have received this email by mistake and delete this email from your system. Please note that any views or opinions presented in this email are solely those of the author and do not necessarily represent those of the organisation. Finally, the recipient should check this email and any attachments for the presence of viruses. The organisation accepts no liability for any damage caused by any virus transmitted by this email. -------------- next part -------------- An HTML attachment was scrubbed... URL: http://bestpractical.com/pipermail/rt-devel/attachments/20050128/b4f26ead/attachment.htm From stefan at van-boxmer.de Fri Jan 28 02:17:21 2005 From: stefan at van-boxmer.de (Stefan Fischer) Date: Fri Jan 28 04:31:34 2005 Subject: [Rt-devel] RT 3.2.2 custom scrip with user defined condition fails In-Reply-To: <016701c50496$86cb19b0$1e0d250a@armdev.am> References: <1515E176F4AD2E4AA41D92D7C006BC47015A9C59@ALHM4SRV.alhpridom.as> <41F91F0B.5000800@van-boxmer.de> <016701c50496$86cb19b0$1e0d250a@armdev.am> Message-ID: <41F9E701.1010703@van-boxmer.de> Hi, i have test it but with no luck. The error ist the same as described. A function in /usr/share/perl5/IO/ScalarArray.pm on line 350 generates the errror message. "$/ as given is currently unsupportet" But i don't know why. Greetings - Stefan Eduard Rushanian wrote: > Hi, > > I'm not sure but you can try following: > > Custom Condition: > return undef unless (($self->TransactionObj->Field eq '2') and > ($self->TransactionObj->NewValue eq '20 Antrag eingetroffen')); > > Best, > Eduard > > ----- Original Message ----- > From: "Stefan Fischer" > To: > Sent: Thursday, January 27, 2005 9:04 PM > Subject: [Rt-devel] RT 3.2.2 custom scrip with user defined condition fails > > > >>Hello, >> >>i have asked this problem a view days bevore in the rt-users section, >>but nobody could help me. >> >>I have a custom condition >> >>if (($self->TransactionObj->Type eq "CustomField") and >> ($self->TransactionObj->Field eq 2) and >> ($self->TransactionObj->NewValue eq "20 Antrag eingetroffen")) { >> return(1); >>} else { >> return(undef); >>} > > -- Mit freundlichen Gr??en / Kind Regards __\/__ . / ^ _ \ . |\| (o)(o) |/| _______.OOOo--oo--oOOO._________ | | | Stefan Fischer | | IT-Consulting | | | | Mobil...: +49 163 8027222 | | B?ro....: +49 89 36055848 | | Fax.....: +49 89 36055846 | | e-mail..: stefan@van-boxmer.de | |________________Oooo.___________| .oooO ( ) ( ) ) / \ ( (_/ \_) Question authority. From af at syd.de Fri Jan 28 04:38:16 2005 From: af at syd.de (Alexander Finger) Date: Fri Jan 28 05:12:38 2005 Subject: [Rt-devel] Reports: Customfields In-Reply-To: <41F9E701.1010703@van-boxmer.de> References: <1515E176F4AD2E4AA41D92D7C006BC47015A9C59@ALHM4SRV.alhpridom.as> <41F91F0B.5000800@van-boxmer.de> <016701c50496$86cb19b0$1e0d250a@armdev.am> <41F9E701.1010703@van-boxmer.de> Message-ID: <41FA0808.3060109@syd.de> Hi, I am trying to generate a "simple" report using rt 3.4rc5; I want all tickets which have been created or resolved in a given period of time or are new/open. The condition I put for that is ( (Queue = 'queue0') OR (Queue = 'queue1') OR (Queue = 'queue2') ) AND ((Created > '2005-01-17') AND (Created < '2005-01-24') OR (Resolved > '2005-01-17') AND (Resolved < '2005-01-24') OR ((Status = 'new') OR (Status = 'open')) ) The query itself seems to return what I want. Now, I want as well to show some custom fields in the resulting report: % while (my $ticket = $tix->Next) { % my $TimeToPick; % my $TimeToResolve; <% $ticket->id %> <% $ticket->Created%> <% $ticket->Creator%> .. <% $ticket->CustomField.{'MyCustomfield'} %> % } For the "CustomField", I tried various things, but either I get a mason error straight away or the message that the Function is not implemented. (Ah, and I'm talking of both global and local customfields). If anyone could hint out to me if I have to add another RT:: Module or use another function to show the custom field that would be very much appreciated. cheers Alex -- Alexander Finger callto://ch.eurospot.af mailto:af@syd.de From jesse at bestpractical.com Fri Jan 28 07:21:10 2005 From: jesse at bestpractical.com (Jesse Vincent) Date: Fri Jan 28 07:21:15 2005 Subject: [Rt-devel] Reports: Customfields Message-ID: <20050128122110.GV337@bestpractical.com> > <% $ticket->CustomField.{'MyCustomfield'} %> Assuming those are single value custom fields, I think what you want is: <%$ticket->FirstCustomFieldValue('MyCustomField')%> Best, Jesse -- From michael.fernandez at dms.at Fri Jan 28 07:46:18 2005 From: michael.fernandez at dms.at (=?ISO-8859-1?Q?Micha=EBl_Fernandez?=) Date: Fri Jan 28 07:48:05 2005 Subject: [Rt-devel] Porting RT on MaxDB Message-ID: <41FA341A.1010408@dms.at> Hi, I'm working on porting RT on the database MaxDB (the newest version of SapDB). Simple functionalities of RT are already working. Nevertheless, the table "Transactions", necessary for RT system, is a system table in MaxDB. Will you be ready to change this table name in new release of RT, allowing porting on MaxDB ? Thanks for your consideration, Best Regards, Michael Fernandez DMS GmbH. From jesse at bestpractical.com Fri Jan 28 07:52:03 2005 From: jesse at bestpractical.com (Jesse Vincent) Date: Fri Jan 28 07:52:08 2005 Subject: [Rt-devel] Re: Porting RT on MaxDB In-Reply-To: <41FA2B0A.1030905@dms.at> References: <41FA2B0A.1030905@dms.at> Message-ID: <20050128125203.GW337@bestpractical.com> On Fri, Jan 28, 2005 at 01:07:38PM +0100, Michal Fernandez wrote: > Hi, > > I'm working on porting RT on the database MaxDB (the newest version of > SapDB). > > Simple functionalities of RT are already working. > > Nevertheless, the table "Transactions", necessary for RT system, is a > system table in MaxDB. > > Will you be ready to change this table name in new release of RT, > allowing porting on MaxDB ? I'm quite leery of changing one of RT's core table names to support a single database, especially one with such small marketshare. Is there no way to tell MaxDB to use a different tablespace for system and user tables? > Thanks for your consideration, > > Best Regards, > > Michael Fernandez > DMS GmbH. > -- From jesse at bestpractical.com Fri Jan 28 09:56:44 2005 From: jesse at bestpractical.com (Jesse Vincent) Date: Fri Jan 28 09:56:51 2005 Subject: [Rt-devel] Configuration Extraction Tool In-Reply-To: <20050125110450.550.qmail@hm28.locaweb.com.br> References: <20050124203124.GD24460@bestpractical.com> <20050125110450.550.qmail@hm28.locaweb.com.br> Message-ID: <20050128145643.GC337@bestpractical.com> On Tue, Jan 25, 2005 at 09:04:50AM -0200, Lucas Fernando Rosada wrote: > Thanks, Jesse. > I have another doubt: does the factory script (rt-path/sbin/factory) supports Postgres database? When I tried to execute it with a PG database, it crashed when the DESCRIBE commanda was tried (my $sth = $dbh->prepare("DESCRIBE $table")). Factory is mostly designed to run against mysql when we're first generating RT's schema. IIRC, RTFM has a version that works on postgres. > Thanks a lot, > > Lucas F. Rosada > lucas@dextra.com.br > > On Mon, 24 Jan 2005 15:31:24 -0500, Jesse Vincent escreveu: > > > De: Jesse Vincent > > Data: Mon, 24 Jan 2005 15:31:24 -0500 > > Para: Lucas Fernando Rosada > > Assunto: Re: [Rt-devel] Configuration Extraction Tool > > > > > > On Mon, Jan 24, 2005 at 11:22:12AM -0200, Lucas Fernando Rosada wrote: > > > > > > Hello, does anyone know any RT tool that generate an initialdata file from the data that's into the database? > > > > > > > rt-dump-database in RT 3.4 > > > > Jesse > > > > > Thaks, > > > > > > Lucas F. Rosada > > > lucas@dextra.com.br > > > _______________________________________________ > > > Rt-devel mailing list > > > Rt-devel@lists.bestpractical.com > > > http://lists.bestpractical.com/cgi-bin/mailman/listinfo/rt-devel > > > > > > > -- > > > > > > > -- From jesse at bestpractical.com Fri Jan 28 10:35:56 2005 From: jesse at bestpractical.com (Jesse Vincent) Date: Fri Jan 28 10:36:00 2005 Subject: [Rt-devel] [PATCH] DBIx::SB tests generalization In-Reply-To: <41F4017A.80506@acronis.com> References: <41F4017A.80506@acronis.com> Message-ID: <20050128153556.GO337@bestpractical.com> Thanks. Applied. (Except for the bit that tests code you haven't submitted yet) On Sun, Jan 23, 2005 at 10:56:42PM +0300, Ruslan U. Zakirov wrote: > Hello. > Attached patch is ready to apply, but part of the work to make SB tests > more generic. > > t/01basic.t - introduces some basic tests, like different handle loading > and testing that constructor return right objects. Now you don't need to > duplicate that tests in other test files. > > t/utils.pl - is library of utils required for tests. Now it has: > @SupportedDrivers - list of supported back ends > get_handle - function that returns new handle by name > > Regards. Ruslan. > > _______________________________________________ > Rt-devel mailing list > Rt-devel@lists.bestpractical.com > http://lists.bestpractical.com/cgi-bin/mailman/listinfo/rt-devel -- From jesse at bestpractical.com Fri Jan 28 10:36:07 2005 From: jesse at bestpractical.com (Jesse Vincent) Date: Fri Jan 28 10:36:15 2005 Subject: [Rt-devel] [PATCH] Quote identifiers. In-Reply-To: <41F2D9A6.8070602@acronis.com> References: <41F2A19E.3050808@acronis.com> <20050122190125.GS24460@bestpractical.com> <41F2A3A1.1000701@acronis.com> <20050122191404.GT24460@bestpractical.com> <41F2A8F7.9020907@acronis.com> <20050122200313.GW24460@bestpractical.com> <41F2D9A6.8070602@acronis.com> Message-ID: <20050128153607.GP337@bestpractical.com> Thanks. Applied. On Sun, Jan 23, 2005 at 01:54:30AM +0300, Ruslan U. Zakirov wrote: > Jesse Vincent wrote: > > > >>Speed. See next benchmark > > > > > >At the expense of clarity. I hate perl's method dispatch > >performance. Will you update the patch with a comment about _why_ > >you're doing something that looks like it should be optimized away so > >that someone else doesn't "fix" it? > I resubmit patch where thing looks like you want. Done more complex test > that insert 5000 records into SQLite DB. Same time results for both > variants. > > > >>#!/usr/bin/perl -w > >> > >> > >> > >>use strict; > >>use Benchmark qw(cmpthese); > >>my $obj = new Dummy; > >> > >> > >> > >>cmpthese (-6, { > >> 'with $dbh' => sub { $obj->w_cache(); }, > >> 'without $dbh' => sub { $obj->wo_cache() }, > >> }); > >> > >> > >> > >>package Dummy; > >> > >> > >> > >>sub new { return bless { dbh => { foo => 'bar' } }, 'Dummy' } > >>sub dbh { my $self = shift; return $self->{'dbh'} } > >> > >> > >> > >>sub w_cache { > >> my $self = shift; > >> my $dbh = $self->dbh; > >> for(1..5) { my $x = $dbh->{foo} } > >>} > >>sub wo_cache { > >> my $self = shift; > >> for(1..5) {my $x = $self->dbh->{foo} } > >>} > >>Results: > >> Rate without $dbh with $dbh > >>without $dbh 43465/s -- -40% > >>with $dbh 72127/s 66% -- > >> > >>Something similar happens in sub Insert I call quote_identifier in map > >>context for each column, also dbh sub in SB::Handle more complex then in > >>benchmark. > >> > >>Similar situation in UpdateRecordValue. > >> > > > > > -- From jesse at bestpractical.com Fri Jan 28 11:13:39 2005 From: jesse at bestpractical.com (Jesse Vincent) Date: Fri Jan 28 11:13:44 2005 Subject: Fw: [Rt-devel] Native Japanese Support In-Reply-To: <01a701c50445$47318790$9900a8c0@hfwsrv1> References: <01a701c50445$47318790$9900a8c0@hfwsrv1> Message-ID: <20050128161338.GE337@bestpractical.com> > > We are using RT 3.2.2, Perl v5.8.3 and apache 2.0.51 on a fedora core 2 box > > The system response Japanese with messed code (my gut feeling is that it > > cannot detect what is the origional code) That should be easy to check by looking at the characterset conversion headers in the mail RT stores. Jesse From chris at meta4-group.com Fri Jan 28 11:13:00 2005 From: chris at meta4-group.com (Chris Li) Date: Fri Jan 28 12:01:21 2005 Subject: Fw: [Rt-devel] Native Japanese Support References: <01a701c50445$47318790$9900a8c0@hfwsrv1> <20050128161338.GE337@bestpractical.com> Message-ID: <001801c50554$3dfb2c70$b103a8c0@chrisli> > > > >> > We are using RT 3.2.2, Perl v5.8.3 and apache 2.0.51 on a fedora core 2 >> > box >> > The system response Japanese with messed code (my gut feeling is that >> > it >> > cannot detect what is the origional code) > > > That should be easy to check by looking at the characterset conversion > headers in the mail RT stores. > > Jesse The problem always happen when the incoming email is a shift-JIS one, and also when sending out email, the email sent is always UNICODE, which are definitely lack of support among the Japanese.. I would be greatful if I can change the outgoing email to a EUC / shiftJIS one :-) Chris From jesse at bestpractical.com Fri Jan 28 12:10:13 2005 From: jesse at bestpractical.com (Jesse Vincent) Date: Fri Jan 28 12:10:19 2005 Subject: Fw: [Rt-devel] Native Japanese Support In-Reply-To: <001801c50554$3dfb2c70$b103a8c0@chrisli> References: <01a701c50445$47318790$9900a8c0@hfwsrv1> <20050128161338.GE337@bestpractical.com> <001801c50554$3dfb2c70$b103a8c0@chrisli> Message-ID: <20050128171013.GN337@bestpractical.com> On Sat, Jan 29, 2005 at 12:13:00AM +0800, Chris Li wrote: > > > > > > > >>> We are using RT 3.2.2, Perl v5.8.3 and apache 2.0.51 on a fedora core 2 > >>> box > >>> The system response Japanese with messed code (my gut feeling is that > >>> it > >>> cannot detect what is the origional code) > > > > > >That should be easy to check by looking at the characterset conversion > >headers in the mail RT stores. > > > >Jesse > > The problem always happen when the incoming email is a shift-JIS one, and > also when sending out email, the email sent is always UNICODE, which are > definitely lack of support among the Japanese.. > > I would be greatful if I can change the outgoing email to a EUC / shiftJIS > one :-) You can. See the configuration file. > > Chris > > -- From dimitry.faynerman at hypermediasystems.com Fri Jan 28 12:21:14 2005 From: dimitry.faynerman at hypermediasystems.com (Dimitry Faynerman) Date: Fri Jan 28 12:37:44 2005 Subject: Fw: [Rt-devel] Native Japanese Support In-Reply-To: <001801c50554$3dfb2c70$b103a8c0@chrisli> Message-ID: <000001c5055d$c883deb0$1b04140a@DimitryFaynerman> I'm using Japanese language in RT (actually in 6 or 7 different RTs we have here), versions 3.0.9 and 3.2.1 I've setup @EmailInputEncodings = qw(utf-8 iso-8859-1 us-ascii) Set($EmailOutputEncoding , 'shift-jis') in my RT_SiteConfig.pm file and it works great From jesse at bestpractical.com Fri Jan 28 13:47:14 2005 From: jesse at bestpractical.com (Jesse Vincent) Date: Fri Jan 28 13:47:19 2005 Subject: [Rt-devel] Re: [rt-users] RT becoming unstable / unable to add attachments (requires apache restart to fix?) In-Reply-To: <573F1E2597A7974E924A5F814CAC166D0DF27E47@da0shme010.match.corp> References: <573F1E2597A7974E924A5F814CAC166D0DF27E47@da0shme010.match.corp> Message-ID: <20050128184714.GV337@bestpractical.com> On Fri, Jan 28, 2005 at 12:35:05PM -0600, Rodney Caston wrote: > Do you suggest mod_perl 1.0 or 2.0 ? Personally, I like FastCGI, mod_perl1 and mod_perl 2.0, in that order. Note also that mod_perl 2.0 isn't out yet. Please do try to cc rt-devel when replying, rather than sending mail directly to someone on the list helping you out. That way, everything ends up in the archives. Best, Jesse From af at syd.de Sat Jan 29 06:16:41 2005 From: af at syd.de (Alexander Finger) Date: Sat Jan 29 06:38:38 2005 Subject: [Rt-devel] Reports: Customfields In-Reply-To: <200501281647.j0SGlUY8022320@edison.serreyn.com> References: <200501281647.j0SGlUY8022320@edison.serreyn.com> Message-ID: <41FB7099.1020601@syd.de> Hi, $ticket->FirstCustomField('fieldname') does the trick, but only _once_. We will need a number of custom fields in the report, and the result I get is TicketID CF1 CF2 CF3 1 A A A 2 B B e.g. once the function returns the FirstCustomField Value, it always returns the value it returned on the first time. Is it a feature (e.g. my fault) or a bug? rgds af From stefan at van-boxmer.de Sat Jan 29 15:09:53 2005 From: stefan at van-boxmer.de (Stefan Fischer) Date: Sat Jan 29 15:14:48 2005 Subject: [Rt-devel] bug? in RT 3.2.2 custom scrip with user defined condition fails In-Reply-To: <41F9E701.1010703@van-boxmer.de> References: <1515E176F4AD2E4AA41D92D7C006BC47015A9C59@ALHM4SRV.alhpridom.as> <41F91F0B.5000800@van-boxmer.de> <016701c50496$86cb19b0$1e0d250a@armdev.am> <41F9E701.1010703@van-boxmer.de> Message-ID: <41FBED91.6040308@van-boxmer.de> Hi, i am able to reproduce the problem on a fresh installation on another debian machine with RT 3.2.2. Creating a scrip with a custom condition like return undef unless (($self->TransactionObj->Field eq '2') and ($self->TransactionObj->NewValue eq '20 Antrag eingetroffen')); with an action like "reply to requestor with global template xy" fails with this error in syslog on modul /usr/share/perl5/IO/ScalarArray.pm line 350. Jan 29 21:00:30 debian RT: #1/18 - Scrip 15 AutoMsg (/usr/share/request-tracker3.2/lib/RT/Action/SendEmail.pm:231) Jan 29 21:00:30 debian RT: Could not send mail. -$/ as given is currently unsupported at /usr/share/perl5/MIME/Decoder/NBit.pm line 140 Stack: [/usr/share/perl/5.8/Carp.pm:86] [/usr/share/perl5/IO/ScalarArray.pm:350] [/usr/share/perl5/MIME/Decoder/NBit.pm:140] [/usr/share/perl5/MIME/Decoder.pm:263] [/usr/share/perl5/MIME/Entity.pm:1856] [/usr/share/perl5/MIME/Entity.pm:1830] [/usr/share/perl5/MIME/Entity.pm:1750] [/usr/share/perl5/MIME/Entity.pm:1886] [/usr/share/perl5/MIME/Entity.pm:1889] [/usr/share/request-tracker3.2/lib/RT/Action/SendEmail.pm:251] [/usr/share/request-tracker3.2/lib/RT/Action/SendEmail.pm:102] [/usr/share/request-tracker3.2/lib/RT/ScripAction_Overlay.pm:225] [/usr/share/request-tracker3.2/lib/RT/Scrip_Overlay.pm:497] [/usr/share/request-tracker3.2/lib/RT/Scrips_Overlay.pm:188] [/usr/share/request-tracker3.2/lib/RT/Transaction_Overlay.pm:154] [/usr/share/request-tracker3.2/lib/RT/Ticket_Overlay.pm:3707] Is there another one who can reconstruct this error condition? good byte! Stefan Stefan Fischer schrieb: > Hi, > > i have test it but with no luck. The error ist the same as described. A > function in /usr/share/perl5/IO/ScalarArray.pm on line 350 generates the > errror message. "$/ as given is currently unsupportet" But i don't know > why. > > Greetings > - Stefan > > Eduard Rushanian wrote: > >> Hi, >> >> I'm not sure but you can try following: >> >> Custom Condition: >> return undef unless (($self->TransactionObj->Field eq '2') and >> ($self->TransactionObj->NewValue eq '20 Antrag eingetroffen')); >> >> Best, >> Eduard >> >> ----- Original Message ----- From: "Stefan Fischer" >> >> To: >> Sent: Thursday, January 27, 2005 9:04 PM >> Subject: [Rt-devel] RT 3.2.2 custom scrip with user defined condition >> fails >> >> >> >>> Hello, >>> >>> i have asked this problem a view days bevore in the rt-users section, >>> but nobody could help me. >>> >>> I have a custom condition >>> >>> if (($self->TransactionObj->Type eq "CustomField") and >>> ($self->TransactionObj->Field eq 2) and >>> ($self->TransactionObj->NewValue eq "20 Antrag eingetroffen")) { >>> return(1); >>> } else { >>> return(undef); >>> } >> >> >> > > - From Ruslan.Zakirov at acronis.com Sun Jan 30 01:18:15 2005 From: Ruslan.Zakirov at acronis.com (Ruslan U. Zakirov) Date: Sun Jan 30 01:23:22 2005 Subject: [Rt-devel] Reports: Customfields In-Reply-To: <41FB7099.1020601@syd.de> References: <200501281647.j0SGlUY8022320@edison.serreyn.com> <41FB7099.1020601@syd.de> Message-ID: <41FC7C27.4060907@acronis.com> See http://wiki.bestpractical.com/index.cgi?AddCustomFieldsValuesToMail Alexander Finger wrote: > Hi, > > $ticket->FirstCustomField('fieldname') does the trick, but only _once_. > We will need a number of custom fields in the report, and the result I > get is > > TicketID CF1 CF2 CF3 > 1 A A A > 2 B B > > e.g. once the function returns the FirstCustomField Value, it always > returns the value it returned on the first time. > > Is it a feature (e.g. my fault) or a bug? > > > rgds > af > > _______________________________________________ > Rt-devel mailing list > Rt-devel@lists.bestpractical.com > http://lists.bestpractical.com/cgi-bin/mailman/listinfo/rt-devel From jesse at bestpractical.com Sun Jan 30 12:48:06 2005 From: jesse at bestpractical.com (Jesse Vincent) Date: Sun Jan 30 12:48:12 2005 Subject: [Rt-devel] RT 3.2.3rc2 Message-ID: <20050130174806.GO337@bestpractical.com> I'm pleased to announce the release of the second release candidate for RT 3.2.2, which is likely to be the final RT 3.2 release before RT 3.4.0. Unless we hear about regressions in 3.2.3rc2, this version will be released as RT 3.2.3 this Wednesday. Jesse -- From jesse at bestpractical.com Sun Jan 30 13:00:36 2005 From: jesse at bestpractical.com (Jesse Vincent) Date: Sun Jan 30 13:00:41 2005 Subject: [Rt-devel] bug? in RT 3.2.2 custom scrip with user defined condition fails In-Reply-To: <41FBED91.6040308@van-boxmer.de> References: <1515E176F4AD2E4AA41D92D7C006BC47015A9C59@ALHM4SRV.alhpridom.as> <41F91F0B.5000800@van-boxmer.de> <016701c50496$86cb19b0$1e0d250a@armdev.am> <41F9E701.1010703@van-boxmer.de> <41FBED91.6040308@van-boxmer.de> Message-ID: <20050130180036.GU337@bestpractical.com> > 15 AutoMsg (/usr/share/request-tracker3.2/lib/RT/Action/SendEmail.pm:231) > Jan 29 21:00:30 debian RT: > Could not send > mail. -$/ as given is currently unsupported at > /usr/share/perl5/MIME/Decoder/NBit.pm line 140 Stack: > [/usr/share/perl/5.8/Carp.pm:86] > [/usr/share/perl5/IO/ScalarArray.pm:350] > [/usr/share/perl5/MIME/Decoder/NBit.pm:140] > [/usr/share/perl5/MIME/Decoder.pm:263] > [/usr/share/perl5/MIME/Entity.pm:1856] > [/usr/share/perl5/MIME/Entity.pm:1830] > [/usr/share/perl5/MIME/Entity.pm:1750] > [/usr/share/perl5/MIME/Entity.pm:1886] > [/usr/share/perl5/MIME/Entity.pm:1889] I've seen this error reported a few times of late, but never run into it myself. It would be great if someone could get a small, reliable testcase together for it. Jesse From chris at meta4-group.com Sun Jan 30 23:25:39 2005 From: chris at meta4-group.com (Chris Li) Date: Sun Jan 30 23:29:16 2005 Subject: Fw: [Rt-devel] Native Japanese Support References: <01a701c50445$47318790$9900a8c0@hfwsrv1> <20050128161338.GE337@bestpractical.com> <001801c50554$3dfb2c70$b103a8c0@chrisli> <20050128171013.GN337@bestpractical.com> Message-ID: <005d01c5074c$ebfde0c0$9900a8c0@hfwsrv1> ----- Original Message ----- From: "Jesse Vincent" To: "Chris Li" Cc: "Jesse Vincent" ; Sent: Saturday, January 29, 2005 1:10 AM Subject: Re: Fw: [Rt-devel] Native Japanese Support > > > > On Sat, Jan 29, 2005 at 12:13:00AM +0800, Chris Li wrote: > > > > > > > > > > > >>> We are using RT 3.2.2, Perl v5.8.3 and apache 2.0.51 on a fedora core 2 > > >>> box > > >>> The system response Japanese with messed code (my gut feeling is that > > >>> it > > >>> cannot detect what is the origional code) > > > > > > > > >That should be easy to check by looking at the characterset conversion > > >headers in the mail RT stores. > > > > > >Jesse > > > > The problem always happen when the incoming email is a shift-JIS one, and > > also when sending out email, the email sent is always UNICODE, which are > > definitely lack of support among the Japanese.. > > > > I would be greatful if I can change the outgoing email to a EUC / shiftJIS > > one :-) > > You can. See the configuration file. I just found the parameter.. but however, it only make the change to the email body but not to the subject, any method for that ? > > > > > Chris > > > > > > -- From rmreis at cpqd.com.br Mon Jan 31 04:40:53 2005 From: rmreis at cpqd.com.br (Rodolfo de Moraes Reis) Date: Mon Jan 31 05:30:28 2005 Subject: [Rt-devel] Custom Fields: Validation Dates! Message-ID: <097013530AD6F24F8D21F7CABCB26A250110C023@MAILSRV1.aquarius.cpqd.com.br> Hello There!! I have a custom field that will be used to get dates (dd/mm/aaaa), and I made a javascript code that does the validation entry, but I don't know where can I put this code or if has a scrip solution to my case... I think that I need change something in /Tickets/Elements/EditCustomField but I don't know the exactly point.... Someone could help me??? Thank's for regards. Rodolfo de Moraes Reis CPqD Telecom & IT Solutions Tel.: +55 19 3705-5986 Fax: +55 19 3705-6786 rmreis@cpqd.com.br www.cpqd.com.br From fjmartinez at csi.uned.es Mon Jan 31 08:20:12 2005 From: fjmartinez at csi.uned.es (Francisco Javier =?iso-8859-1?Q?Mart=EDnez?= Martinez) Date: Mon Jan 31 09:35:14 2005 Subject: [Rt-devel] How to make LDAP authentication in RT3 Message-ID: <6.1.2.0.2.20050131141327.04299ce8@csipop.uned.es> Hello folks. Could anyone post the procedure, file, ... to enable LDAP authentication against an external LDAP^server, with RT3 in linux. I know that is possible to authenticate users against an external LDAP server, I had been googling and searching in mail-lists, and I had found a lot of different references, too much references, but not very clear, and most of them refering to RT2, I wonder that it could be the same for RT3. This is the mainly reason of my request. Thanks in advance. From prouss at cae.com Mon Jan 31 09:43:18 2005 From: prouss at cae.com (Patrick Rousseau-Mathieu) Date: Mon Jan 31 09:49:09 2005 Subject: [Rt-devel] How to make LDAP authentication in RT3 Message-ID: Hi, You have to install apache mod_auth et mod_auth_ldap etc... and there you all the code here that override the login connexion. http://www.usit.uio.no/it/rt/modifications.html Patrick. -----Original Message----- From: rt-devel-bounces@lists.bestpractical.com [mailto:rt-devel-bounces@lists.bestpractical.com]On Behalf Of Francisco Javier Mart?nez Martinez Sent: Monday, January 31, 2005 8:20 AM To: rt-users@lists.bestpractical.com Cc: rt-devel@lists.bestpractical.com Subject: [Rt-devel] How to make LDAP authentication in RT3 Hello folks. Could anyone post the procedure, file, ... to enable LDAP authentication against an external LDAP^server, with RT3 in linux. I know that is possible to authenticate users against an external LDAP server, I had been googling and searching in mail-lists, and I had found a lot of different references, too much references, but not very clear, and most of them refering to RT2, I wonder that it could be the same for RT3. This is the mainly reason of my request. Thanks in advance. _______________________________________________ Rt-devel mailing list Rt-devel@lists.bestpractical.com http://lists.bestpractical.com/cgi-bin/mailman/listinfo/rt-devel From af at syd.de Mon Jan 31 10:56:57 2005 From: af at syd.de (Alexander Finger) Date: Mon Jan 31 10:02:17 2005 Subject: [Rt-devel] How to make LDAP authentication in RT3 In-Reply-To: References: Message-ID: <41FE5549.9020004@syd.de> > Could anyone post the procedure, file, ... to enable LDAP authentication > against an external LDAP^server, with RT3 in linux. > > I know that is possible to authenticate users against an external LDAP > server, I had been googling and searching in mail-lists, and I had found a > lot of different references, too much references, but not very clear, and > most of them refering to RT2, I wonder that it could be the same for RT3. > This is the mainly reason of my request. I switched it on on my testbox Accept REMOTE_USER as authenticated: in RT_SiteConfig.pm: Set($WebExternalAuth , 1); #Set($WebFallbackToInternalAuth , 1); #If the user does not exist, create him: Set($WebExternalAuto , 1); // And then in httpd.conf: AuthName "Please type your [firstname.lastname] and your [mail-passwd] to acc ess RT" AuthType Basic AuthLDAPURL ldap://ldap.server.com:389/o=my-company?login?sub?(mail= *) require valid-user mod_auth_ldap is required, of course. The user can then type his "login" attribute to login and can only login if a mail-attribute is present and filled in his user data. You could set this to "accessrt=yes" to make sure only people with an "accessrt"-Attribute of "yes" could access rt. You'd have to manage that attribute, though (create, maintain..). I did not do any modifications on rt itself. cheers Alex -- Alexander Finger callto://ch.eurospot.af mailto:af@syd.de From rkagan at yorku.ca Mon Jan 31 11:46:22 2005 From: rkagan at yorku.ca (Ramon Kagan) Date: Mon Jan 31 12:05:47 2005 Subject: [Rt-devel] 3.4.0rc6 small bug Message-ID: HI, Not really a bug but will be in the future. If you display a ticket and then select the basic modify, the logs show that the CustomField is using the deprecated call. Might want to fix this to remove the chance of a bug later. [Mon Jan 31 16:42:57 2005] [warning]: Queue->CustomFields is deprecated, use Queue->TicketCustomFields instead at /opt/rt3/lib/RT/Queue_Overlay.pm line 507. (/opt/rt3/lib/RT.pm:277) Ramon Kagan York University, Computing and Network Services Information Security - Senior Information Security Analyst (416)736-2100 #20263 rkagan@yorku.ca ----------------------------------- ------------------------------------ I have not failed. I have just I don't know the secret to success, found 10,000 ways that don't work. but the secret to failure is trying to please everybody. - Thomas Edison - Bill Cosby ----------------------------------- ------------------------------------