Re: Postgres HA

On Fri, 5 Jan 2018 13:07:10 -0600
Azimuddin Mohammed <azimeiu(at)gmail(dot)com> wrote:

> Hello,
> I am little confused with how HA works in postgres. Reading the article
> which state as below "*If the primary server fails and the standby server
> becomes the new primary, and then the old primary restarts, you must have a
> mechanism for informing the old primary that it is no longer the primary.
> This is sometimes known as STONITH (Shoot The Other Node In The Head),
> which is necessary to avoid situations where both systems think they are
> the primary, which will lead to confusion and ultimately data loss.*
>
> *Many failover systems use just two systems, the primary and the standby,
> connected by some kind of heartbeat mechanism to continually verify the
> connectivity between the two and the viability of the primary. It is also
> possible to use a third system (called a witness server) to prevent some
> cases of inappropriate failover, but the additional complexity might not be
> worthwhile unless it is set up with sufficient care and rigorous testing.*
> *PostgreSQL does not provide the system software required to identify a
> failure on the primary and notify the standby database server. Many such
> tools exist and are well integrated with the operating system facilities
> required for successful failover, such as IP address migration."*
>
> Can someone explain how the HA failback will take place

The failback need either to rebuild the old master as a standby (rsync,
pg_basebackup, restore PITR, ...) or to use pg_rewind to rewind the old master
to a point where it can catch up with the new master.

Some tools tries to automate failback using pg_rewind (patroni, repmgr), but I
have no experience with them.

> and what open source tools we can use to make sure once the primary server
> which failed over to slave will mark itself as slave.

There's a lot of open source tools to build some HA around PgSQL: Repmgr,
Patroni (based on etcd or zookeeper), PAF (based on Pacemaker), etc. You will
have to spend a lot of time to make extensive tests, understand them, pick one
and document your cluster.

Regards,