From: | "Scott Otis" <scott(dot)otis(at)intand(dot)com> |
---|---|
To: | <pgsql-performance(at)postgresql(dot)org> |
Subject: | Seeking performance advice and explanation for high I/O on 8.3 |
Date: | 2009-09-02 20:44:42 |
Message-ID: | F406B740738CB44A8BCBEF9734ECF663683EC9@albus.intand.com |
Views: | Raw Message | Whole Thread | Download mbox | Resend email |
Thread: | |
Lists: | pgsql-performance |
Would love to get some advice on how to change my conf settings / setup
to get better I/O performance.
Server Specs:
2x Intel Xeon Quad Core (@2 Ghz - Clovertown,L5335)
4GB RAM
4x Seagate 73GB SAS HDD 10k RPM - in RAID ( stripped and mirrored )
FreeBSD 6.4
Apache 2.2
PostgreSQL 8.3.6
PHP 5.2.9
~1500 databases w/ ~60 tables each
Total I/O (these number are pretty constant throughout the day):
Reads: ~ 100 / sec for about 2.6 Mb/sec
Writes: ~ 400 /sec for about 46.1Mb/sec
Conf settings:
listen_addresses = '*'
max_connections = 600
ssl = on
password_encryption = on
shared_buffers = 1GB
work_mem = 5MB
maintenance_work_mem = 256MB
max_fsm_pages = 2800000
max_fsm_relations = 160000
synchronous_commit = off
checkpoint_segments = 6
checkpoint_warning = 30s
effective_cache_size = 1GB
pg_stat_bgwriter:
checkpoints_timed: 16660
checkpoints_req: 1309
buffers_checkpoint: 656346
buffers_clean: 120922
maxwritten_clean: 1
buffers_backend: 167623
buffers_alloc: 472802349
This server also handles web traffic and PHP script processing.
Most of the SQL happening is selects - very little inserts, updates and
deletes comparatively.
I have noticed that most/all of the I/O activity is coming from the
stats collector and autovacuum processes. Would turning off the stats
collector and autovacuum be helpeful / recommended? Could I change my
checkpoint_* or bgwriter_* conf values to help?
Let me know if you need more information / stats.
Any help would be much appreciated.
Thanks,
Scott Otis
CIO / Lead Developer
Intand
From | Date | Subject | |
---|---|---|---|
Next Message | Magnus Hagander | 2009-09-03 07:51:49 | Re: Help: how to speed up query after db server reboot |
Previous Message | Greg Jaman | 2009-09-02 20:31:29 | Re: partition query using Seq Scan even when index is present |