Re: Disable Streaming Replication without restarting either master or slave

I understand the way things work currently, but would it not be more
convenient to let the replication be stopped and started from SQL commands
like:

On Master:

select pg_pause_streaming_replication(slave_fqdn);
select pg_resume_streaming_replication(slave_fqdn);

On Slave:

select pg_pause_streaming_replication();
select pg_pause_streaming_replication();

If Cascaded Streaming Replication is confugured, then

On Slave:

select pg_start_streaming_replication(cascaded_slave_fqdn);
select pg_stop_streaming_replication(cascaded_slave_fqdn);

Or an OS executable [ either in the core or a contrib module] like

pg_replication -p primary_fqdn -s slave_fqdn -a pause
pg_replication -p primary_fqdn -s slave_fqdn -a resume
pg_replication -p primary_fqdn -s slave_fqdn -a setup

[ the last one would awesome; and i do not think that it is impossible, and
would be loved by one and all ]

Regards,
Samba

========================================================

On Wed, May 30, 2012 at 12:07 AM, Fujii Masao <masao(dot)fujii(at)gmail(dot)com> wrote:

> On Wed, May 30, 2012 at 2:38 AM, Michael Nolan <htfoot(at)gmail(dot)com> wrote:
> >
> >
> > ---------- Forwarded message ----------
> > From: Michael Nolan <htfoot(at)gmail(dot)com>
> > Date: Tue, May 29, 2012 at 1:37 PM
> > Subject: Re: [GENERAL] Disable Streaming Replication without restarting
> > either master or slave
> > To: Fujii Masao <masao(dot)fujii(at)gmail(dot)com>
> >
> >
> >
> >
> > On Tue, May 29, 2012 at 1:15 PM, Fujii Masao <masao(dot)fujii(at)gmail(dot)com>
> wrote:
> >>
> >> On Tue, May 29, 2012 at 10:17 PM, François Beausoleil
> >> <francois(at)teksol(dot)info> wrote:
> >
> >
> >>
> >> > Are per-chance looking for pg_xlog_replay_pause() and
> >> > pg_xlog_replay_resume() ?
> >>
> >> Those can pause and resume WAL replay in the standby, but not streaming
> >> replication. Even while WAL replay is being paused, WAL can be streamed
> >> from the master to the standby.
> >>
> >> Regards,
> >>
> >> --
> >> Fujii Masao
> >
> >
> > So, that means that the only ways to stop streaming replication are to
> stop
> > the slave server, to disable access to the master via the pg_hba.conf
> file
> > (requiring the master configs be reloaded) or to set the trigger file on
> the
> > slave to tell it to stop replicating the master.
> >
> > And if the master/slave are set to synchronous streaming replication,
> your
> > options are more limited, since the master has to know to stop waiting
> for
> > the synchronous slave to respond.
> >
> > Once the slave has gone out of asynchronous replication mode, wuld it be
> > possible to resume asynchronous replication by stopping the slave server,
> > removing the trigger file, and restarting it in asynchronous streaming
> > replication mode? This would, at a minimum, depend on how many updates
> have
> > occurred on the master during the time streaming replication was disabled
> > and having all the WAL files available, right?
>
> You'd like to restart the *promoted* standby server as the standby again?
> To do this, a fresh base backup must be taken from the master onto
> the standby before restarting it, even if there has been no update since
> the standby had been promoted.
>
> Regards,
>
> --
> Fujii Masao
>
> --
> Sent via pgsql-general mailing list (pgsql-general(at)postgresql(dot)org)
> To make changes to your subscription:
> http://www.postgresql.org/mailpref/pgsql-general
>