Hi all,

Vick Khera <vivek at khera.org> writes:

> So in prep for upgrading from 2.1 to 2.2 this weekend, I upgraded my
> server's OS from FreeBSD 9.1 to 9.2 (a fairly minor update, as OS updates
> go).
>
> Since the upgrade, the slon connected to the replica DB on that upgraded
> server will stop after just about 58 to 59 minutes. Restarting the slon
> daemon allows the replication to continue and fairly quickly catch up.

I have similar problems. It is 45 minutes +/- remote_listen_timeout for me.

[...]

I my case it's "master" on FreeBSD 9.2 and a "slave" on FreeBSD 9.1. It
seems that in some time around net.inet.tcp.keepidle (sysctl -a to
check) when keepalive from OS kicks in the transsmission will be droped
depending on your keepalive configuration. You should also see error
messages in postgresql logs like:

getsockopt(TCP_KEEPCNT) failed: Protocol not available

> Any ideas? This is so confusing because it is such an odd time interval
> before it locks up. What's magical about 58 minutes?

My wild guess is:

http://lists.freebsd.org/pipermail/freebsd-stable/2013-November/075781.html


The problem I have is, that slon on slave reconnects to database on
master but the backend process on master stays forever taking
resournces. From first connection I have like 1 hour "in
production" - everything is just fine, then comes keepalive + 
remote_listen_timeout, drops the connection and slon reconnects every
remote_listen_timeout creating new backend processes.

I have one simple workaround for this: crontab + slon restart every
hour.

Cheers,
thorkill

PS. Postgresql 9.3.2 used on both nodes.
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: application/pgp-signature
Size: 489 bytes
Desc: not available
Url : http://lists.slony.info/pipermail/slony1-general/attachments/20140314/42cc8943/attachment.pgp