pg_auto_failover
pg_auto_failover copied to clipboard
The online monitor replacement is not working properly.
According to this documentation, it seemed possible to perform this operation online without interruptions.
I executed the command pg_autoctl disable monitor --force sequentially on the primary and multiple secondaries.
Then, I ran the command pg_autoctl enable monitor ... in the order of primary first, followed by the secondaries.
While the primary had no issues, all the secondary nodes experienced a restart of the PostgreSQL process.
It does not work online as expected.
Could you advise on how to replace the monitor without restarting the secondary nodes? 😢