Re: [PERFORM] Postgres 8.2 memory weirdness

2008-01-27 Thread Tory M Blue
On Jan 27, 2008 4:08 PM, Greg Smith <[EMAIL PROTECTED]> wrote: > > More segments means more disk space taken up with them and a longer crash > recovery. Those are the downsides; if you can live with those there's no > reason to run at <100 if that works for you. Fine-tuning here isn't > really t

Re: [PERFORM] Postgres 8.2 memory weirdness

2008-01-27 Thread Greg Smith
On Fri, 25 Jan 2008, Tory M Blue wrote: I doubled the checkpoint segments yesterday and have not seen any warnings. Will run with segments of 100 for a while and see how things look.. Anyway to make sure that there is not a number between 50 and 100 that makes more sense? More segments means m

Re: [PERFORM] Postgres 8.2 memory weirdness

2008-01-25 Thread Tory M Blue
On Jan 24, 2008 10:49 AM, Greg Smith <[EMAIL PROTECTED]> wrote: > 8.2.1 has a nasty bug related to statistics collection that causes > performance issues exactly in the kind of heavy update situation you're > in. That's actually why i asked for the exact 8.2 version. You should > plan an upgrade

Re: [PERFORM] Postgres 8.2 memory weirdness

2008-01-24 Thread Greg Smith
On Thu, 24 Jan 2008, Tory M Blue wrote: Postg: 8.2.1fc6 8.2.1 has a nasty bug related to statistics collection that causes performance issues exactly in the kind of heavy update situation you're in. That's actually why i asked for the exact 8.2 version. You should plan an upgrade as soon

Re: [PERFORM] Postgres 8.2 memory weirdness

2008-01-24 Thread Tory M Blue
On Jan 23, 2008 4:31 PM, Greg Smith <[EMAIL PROTECTED]> wrote: > Generally if you have a system doing many updates and inserts that slows > for that long, it's because it hit a checkpoint. I'm not sure what your > memory-related issues are but it's possible that might be from a backlog > of sessi

Re: [PERFORM] Postgres 8.2 memory weirdness

2008-01-23 Thread Greg Smith
On Wed, 23 Jan 2008, Tory M Blue wrote: I have hundreds of thousands of updates, inserts a day. But what I'm seeing is my server appears to "deallocate" memory (for the lack of a better term) and performance goes to heck, slow response, a sub second query takes anywhere from 6-40 seconds to co

[PERFORM] Postgres 8.2 memory weirdness

2008-01-23 Thread Tory M Blue
I'm not sure what is going on but looking for some advice, knowledge. I'm running multiple postgres servers in a slon relationship. I have hundreds of thousands of updates, inserts a day. But what I'm seeing is my server appears to "deallocate" memory (for the lack of a better term) and performanc