[rt-users] Performance on PostgreSQL

Joby Walker joby at u.washington.edu
Fri Oct 13 18:48:16 EDT 2006


OIDs were used to get the id of the last insert.  The was abandoned 
because of pgsql 8.1, but it was a horrible way to get the last id 
anyway.  Current versions of SB insert the row and then get the current 
value of the sequence associated with that table.  This is much faster 
and *should* be safe.

I've patched our SB (1.36) to work like most of the other db object 
abstraction systems I've used -- get the next val of the sequence and 
then use the value in the insert, this will be correct everytime and in 
all setups.  It is also faster than the default SB-1.36.

Joby Walker
C&C SSG, University of Washington


Kenneth Marshall wrote:
> Vivek,
> 
> I was as surprised as you. It may have been the result of migrations
> from older versions of RT to newer versions. We started at 3.2, are
> currently running 3.4.5, and are preparing to migrate to 3.6.x. If
> the OIDs are not needed, I would like to drop them in our migration
> to 3.6.x. This is with an older version of searchbuilder (1.36) on
> the 3.4.5 system. We could not upgrade to a newer version because it
> caused our DB queries to fail. Do you think that it is safe to remove
> the oid column?
> 
> Ken
> _______________________________________________
> http://lists.bestpractical.com/cgi-bin/mailman/listinfo/rt-users
> 
> Community help: http://wiki.bestpractical.com
> Commercial support: sales at bestpractical.com
> 
> 
> Discover RT's hidden secrets with RT Essentials from O'Reilly Media. 
> Buy a copy at http://rtbook.bestpractical.com



More information about the rt-users mailing list