Re: WAL + SSD = slow inserts? - Mailing list pgsql-performance

From Skarsol
Subject Re: WAL + SSD = slow inserts?
Date
Msg-id CAMt8e=FxA=0x8G-BTuV00BNGSz_-zkykrWG465mg0LBmO+sRxA@mail.gmail.com
Whole thread Raw
In response to Re: WAL + SSD = slow inserts?  (desmodemone <desmodemone@gmail.com>)
Responses Re: WAL + SSD = slow inserts?  (Scott Marlowe <scott.marlowe@gmail.com>)
List pgsql-performance
psql (PostgreSQL) 9.2.5
Red Hat Enterprise Linux Server release 6.4 (Santiago)
Linux 2.6.32-358.6.1.el6.x86_64 #1 SMP Fri Mar 29 16:51:51 EDT 2013 x86_64 x86_64 x86_64 GNU/Linux
All relevant filesystems are ext4

Changes from defaults:
max_connections = 500
shared_buffers = 32000MB
temp_buffers = 24MB
work_mem = 1GB
maintenance_work_mem = 5GB
wal_level = archive
wal_buffers = 16MB
checkpoint_completion_target = 0.9
archive_mode = on
archive_command = 'test ! -f /databases/pg_archive/db/%f && cp %p /databases/pg_archive/db/%f'
effective_cache_size = 64000MB
default_statistics_target = 5000
log_checkpoints = on
stats_temp_directory = '/tmp/pgstat'



On Thu, Dec 5, 2013 at 9:06 AM, desmodemone <desmodemone@gmail.com> wrote:
Hello,
             could you please post the postgresql version, the postgresql.conf, the operative system used, the kernel version and the filesystem used ?

Thank you


2013/12/5 Skarsol <skarsol@gmail.com>
I'm trying to increase the speed of inserts in a database that is on a not super fast storage system. I have installed a pair of SSDs and placed pg_xlog on them but am still getting inserts that take up to a second to complete, with .3 seconds being about average. Iostat doesn't show the SSDs stressed at all, and changing synchronous_commit doesn't seem to affect it one way or the other. Where would I look next for what could be causing the delay?


pgsql-performance by date:

Previous
From: desmodemone
Date:
Subject: Re: WAL + SSD = slow inserts?
Next
From: Dave Johansen
Date:
Subject: Recommendations for partitioning?