Horrible PostgreSQL performance with NFS

Sean Chittenden sean at gigave.com
Thu Jan 19 10:05:37 PST 2006

> > > Did you do those "dd" tests with small block sizes (like 1byte:
> > > bs=1), like somebody on one of those lists suggests, too? Then
> > > we could see, if there is a high latency that ruins
> > > everything...
> > 
> > FYI, PostgreSQL does 8kB I/O by default. This can only be changed
> > by modifying a header file.
> That's definitely small in my book (certainly compared to the 1MB
> block size the first responder suggested), knowing that, you should
> definitely do dd tests with 8k blocks since that's the best
> performance you are likely to get.

PostgreSQL can be configured to use more than 32K blocks, but I recall
65K having problems and shouldn't be used in production.  Tom Lane was
less than amused when he found out I'd added those options to the
ports and issued some strong words suggesting against using anything
but 8K blocks simply because that was the only config that'd seen
years of production use.  YMMV.  -sc

Sean Chittenden
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: application/pgp-signature
Size: 155 bytes
Desc: not available
Url : http://lists.freebsd.org/pipermail/freebsd-performance/attachments/20060119/8d4a1cfc/attachment.bin

More information about the freebsd-performance mailing list