Re: BUG #5851: ROHS (read only hot standby) needs to be restarted manually in somecases.

From: "mark" <dvlhntr(at)gmail(dot)com>
To: "'Robert Haas'" <robertmhaas(at)gmail(dot)com>
Cc: <pgsql-bugs(at)postgresql(dot)org>
Subject: Re: BUG #5851: ROHS (read only hot standby) needs to be restarted manually in somecases.
Date: 2011-01-28 18:03:55
Message-ID: 034301cbbf15$bc078920$34169b60$@com
Views: Raw Message | Whole Thread | Download mbox | Resend email
Thread:
Lists: pgsql-bugs

When showing the setting on the slave or master all tcp_keepalive settings
(idle, interval and count) are showing 0;

The config file shows interval and count commented out, but idle in the
config file is set to 2100.

Possible that "show tcp_keepalive_idle;" isn't reporting accurately ? (or a
value that high isn't be accepted?)

I have reloaded configs and still seeing 0's

I assume you would suggest I turn that number down... a lot.

..: Mark

> -----Original Message-----
> From: Robert Haas [mailto:robertmhaas(at)gmail(dot)com]
> Sent: Friday, January 28, 2011 6:48 AM
> To: Mark
> Cc: pgsql-bugs(at)postgresql(dot)org
> Subject: Re: [BUGS] BUG #5851: ROHS (read only hot standby) needs to be
> restarted manually in somecases.
>
> On Wed, Jan 26, 2011 at 8:24 PM, Mark <dvlhntr(at)gmail(dot)com> wrote:
> > getting a break down in streaming rep. my current work around is to
> restart
> > the PG instance on the ROHS. doesn't seem to affect the master any.
> doesn't
> > require a re-rsync of the base to get replication going again. has
> happened
> > with 9.0.2 twice now in a month.
> >
> >
> >
> > 2011-01-26 08:35:42 MST :: (postgres(at)10(dot)80(dot)2(dot)89) LOG:  could not
> receive
> > data
> > from client: Connection reset by peer
> > 2011-01-26 08:35:42 MST :: (postgres(at)10(dot)80(dot)2(dot)89) LOG:  unexpected EOF
> on
> > standby connection
> >
> > this was all I have in the master's log with the level set to debug
> 1, I
> > have reset it to debug 5 and will just wait till it dies again and
> hopefully
> > get a better idea of what is going on. nothing is being logged to the
> > standby.
>
> Maybe a break in network connectivity is leading the master to think
> that the slave is dead, while the slave still thinks it's connected.
> You might need to adjust the TCP keepalive parameters the slave uses
> to connect to the master.
>
> --
> Robert Haas
> EnterpriseDB: http://www.enterprisedb.com
> The Enterprise PostgreSQL Company

In response to

Responses

Browse pgsql-bugs by date

  From Date Subject
Next Message Dylan Alex Simon 2011-01-29 11:46:55 Re: BUG #5665: catalog/schemapg.h: No such file or directory
Previous Message Tom Lane 2011-01-28 17:18:10 Re: BUG #5855: pgstat wait timeout