Re: [PERFORM] RT3.4 query needed a lot more tuning with 9.2 than it did with 8.1

2013-05-14 Thread Christoph Berg
Re: Mark Felder 2013-05-13 > What version of DBIx-SearchBuilder do you have on that server? The > RT guys usually recommend you have the latest possible so RT is > performing the most sane/optimized queries possible for your > database. I honestly don't know if it will make a difference for > you,

Re: [PERFORM] RT3.4 query needed a lot more tuning with 9.2 than it did with 8.1

2013-05-14 Thread Christoph Berg
Re: Tom Lane 2013-05-06 <1583.1367858...@sss.pgh.pa.us> > The newer rowcount estimates are much further away from reality: > > > Unique (cost=1117.67..1118.46 rows=9 width=1115) (actual > > time=82.646..85.695 rows=439 loops=1) > > > Unique (cost=784205.94..796940.08 rows=145533 width=1061)

[PERFORM] RT3.4 query needed a lot more tuning with 9.2 than it did with 8.1

2013-05-06 Thread Christoph Berg
Hi, this is more of a report than a question, because we thought this would be interesting to share. We recently (finally) migrated an Request Tracker 3.4 database running on 8.1.19 to 9.2.4. The queries used by rt3.4 are sometimes weird, but 8.1 coped without too much tuning. The schema looks li