Re: Vacuums on large busy databases

["Jim C. Nasby" <jimn(at)enterprisedb(dot)com>]

On Thu, Sep 14, 2006 at 11:23:01AM -0400, Francisco Reyes wrote:
> My setup:
> Freebsd 6.1
> Postgresql 8.1.4
> Memory: 8GB
> SATA Disks 
> 
> Raid 1 10 spindles (2 as hot spares)
> 500GB disks (16MB buffer), 7200 rpm
> Raid 10
> 
> Raid 2 4 spindles
> 150GB 10K rpm disks
> Raid 10
> 
> shared_buffers = 10000
> temp_buffers = 1500
> work_mem = 32768                # 32MB
> maintenance_work_mem = 524288   # 512MB
> 
> checkpoint_segments = 64
> Just increased to 64 today.. after reading this may help. Was 5 before.
> 
> pg_xlog on second raid (which sees very little activity)

BTW, on some good raid controllers (with battery backup and
write-caching), putting pg_xlog on a seperate partition doesn't really
help, so you might want to try combining everything.

Even if you stay with 2 partitions, I'd cut pg_xlog back to just a
simple mirror.
-- 
Jim Nasby                                    jimn(at)enterprisedb(dot)com
EnterpriseDB      http://enterprisedb.com      512.569.9461 (cell)