Hi everyone,

I ran into a situation last night where sl_log_1 was over 600MB and took
several hours to finish a logswitch.  I would try logswitch_start(), and
then it would tell me a logswitch was already in progress.  Then, I tried
logswitch_finish(), and it would tell me a logswitch was in progress.  When
I looked at pg_stat_activity, there were usually 5 rows with "FETCH 500 FROM
LOG" in my 11-node slony 2.0.6 cluster.

I was wondering if there'd be any way to make that logswitch go faster (or
perhaps more frequently)?  Would anyone know how to minimize the size of
sl_log_*?  I was thinking about doing things like raising sync_interval to
create less entries in sl_log_* -- would that (or some other adjustment) do
the trick?

Thanks for any help!
--Richard
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://lists.slony.info/pipermail/slony1-general/attachments/20110413/a6a3903d/attachment.htm