[PERFORM] seqscan on UNION'ed views

2013-02-27 Thread Dmitry Karasik
Dear all, I have a problem with seqscan I hope you might help me with. Attached is the simple script that reproduces a database and results, which I have tested both on 9.0.4 and 9.3-devel with identical results. I need to have a sort of a time machine, where select statements on tables could

Re: [PERFORM] Server stalls, all CPU 100% system time

2013-02-27 Thread Andre
On 27/02/2013 9:29 AM, Josh Berkus wrote: Andre, Please see the related thread on this list, High CPU usage / load average after upgrading to Ubuntu 12.04. You may be experiencing some of the same issues. General perspective seems to be that kernels 3.0 through 3.4 have serious performance

Re: [PERFORM] seqscan on UNION'ed views

2013-02-27 Thread Tom Lane
Dmitry Karasik dmi...@karasik.eu.org writes: I need to have a sort of a time machine, where select statements on tables could be easily replaced to select statements on tables as they were some time in the past, including all related table. To do so, I used views (see in the script) that

Re: [PERFORM] Estimation question...

2013-02-27 Thread Matt Daw
Quick follow up... I've found that the row estimate in: explain select count(id) from versions where project_id IN (80,115) AND project_id=115; QUERY PLAN

Re: [PERFORM] Are bitmap index scans slow to start?

2013-02-27 Thread Jeff Janes
On Tue, Feb 26, 2013 at 4:33 PM, Carlo Stonebanks stonec.regis...@sympatico.ca wrote: Is each of these write operations just covering a single row? Does this description apply to just one of the many (how many?) databases, so that there are really 14*N concurrent sessions? ** ** **

Re: [PERFORM] Are bitmap index scans slow to start?

2013-02-27 Thread Carlo Stonebanks
pgbouncer is more for making connections line up single-file when the line is moving at a very fast clip, say 0.01 second per turn. If I were trying to make tasks that can each last for hours or days line up and take turns, I don't think pgbouncer would be the way to go. The recommendation

Re: [PERFORM] Server stalls, all CPU 100% system time

2013-02-27 Thread Josh Berkus
Someone commented they think it might be related to this kernel bug: https://lkml.org/lkml/2012/10/9/210 We have some evidence that that is the case. -- Josh Berkus PostgreSQL Experts Inc. http://pgexperts.com -- Sent via pgsql-performance mailing list

Re: [PERFORM] Are bitmap index scans slow to start?

2013-02-27 Thread Carlo Stonebanks
I had thought you were saying that any one ETL procedure into one database used 14 concurrent threads. But really, each ETL procedure is single-threaded, and there can be up to 5 (or theoretically up to 14) of them running at a time into different databases? Sorry, just caught this.