"Dan Austin" <DAustin@au.westfield.com> writes:
> One of Sun's kernel engineers has come up with a solution that seems to
> have fixed the problem. Instead of increasing work_mem (which we still
> might do) we've mounted a swapfs partition over the pg temp directory:
> # mount | grep pgsql_tmp
> /export/data/pgsql/data/base/64920741/pgsql_tmp on swap [...]
> This has actually reduced the disk activity considerably. So far, the
> blocking problem seems to have gone.
OK, so the stats collector is off the hook: it's not writing anything
in that directory. I'd say you have a pretty conventional
query-optimization problem here after all --- you want to cut the
traffic into those temp files. You ought to be looking at EXPLAIN
ANALYZE info for your common queries.
regards, tom lane
