On Tue, Nov 1, 2022 at 7:26 PM Laurenz Albe <laurenz.albe@xxxxxxxxxxx> wrote:
On Tue, 2022-11-01 at 08:49 -0500, Ron wrote:
> Primary has crashed, according to OP; it's not coming back any time soon. And thus you promote
> Secondary to be New Primary, and go about your work. When the Old Primary comes back up
> (hours or days later), you do a pg_basebackup to make it the New Secondary.
There was no word about failover or "not coming back any time soon" (that's why I asked for
clarification in my answer). In fact, PostgreSQL will often undergo crash recovery and come
back up automatically. See the parameter "restart_after_crash", which defaults to "on".
Sorry, It was my mistake not to mention the failover to the Standby required.
I mean when the Primary Server crashes due to a H/W failure or whatever other reason, Then in my case I am forced to promote the Standby server as Primary until the old primary server H/W vendor replaces the hardware component( it may vary from hours to days as pre SLA terms with H/W vendor).
In this scenario any transactions happened in crashed Primary server which were not transferred due to the delay in transfer to the Standby server. But I am in need to promote the Standby to Primary and it will run for another one day or two, until my Old crashed server failed component is replaced with a new H/W part.
So there will be an inconsistency in the data between the current Primary ( promoted standby) and my recovered /repaired old server ( which was primary before the crash).
How to handle this data inconsistency or what best mechanism needs to be implemented if you were me in this use case scenario?
Yours,
Laurenz Albe
