[rt-users] Question about ticket bulk update

[Gary Mason]

HI,

I have RT 3.6.7 running on Postgresql 8.3

General performance was very poor, so we did some optimising last week.  
We added the following indices :-

rtdb=# create index groups2_lower on groups(lower(type),instance,lower(domain));
rtdb=# create index users1_lower on users(lower(name));
rtdb=# create index users4_lower on users(lower(emailaddress));
rtdb=# create index groupmembers_groupid on groupmembers(groupid);
rtdb=# create index cachedgroupmembers_immediateparentid on cachedgroupmembers(immediateparentid);
rtdb=# create index cachedgroupmembers_immediateparentid on cachedgroupmembers(immediateparentid);

The overall improvement in performance is staggering.  Happy users.

However, it appears that these actions may have broken the ticket bulk update routine.

When a user clicks on the ticket bulk update link now, they get a screen with the tickets on and then a very long wait before the list of users is given to them.  Before last week, this list just contained our local users, but now the list contains every single person who has every emailed in a ticket as well, which is around a million.

Firstly, did these extra indices cause this problem, and secondly, what can I do to get this list back to just our internal user list ?

Thanks,
Gary