Re: [HACKERS] Patch: Implement failover on libpq connect level.

From: pgsql-jdbc-owner(at)postgresql(dot)org
> [mailto:pgsql-jdbc-owner(at)postgresql(dot)org] On Behalf Of Craig Ringer
> >> If true, that's insane. That can be different on each connection to
> >> the cluster and can change tens of thousands of times per second on
> >> any connection!
> >
> > I don't think that's insane. The command is only being issued at
> > connection startup, and will only be different on different
> > connections if it's been configured that way.
>
> I agree. However, I also think we should make sure add a GUC_REPORT var
> in v10 that lets a client tell whether it's connected to a read/write or
> read-only node right from the start. This will save an unnecessary
> round-trip and some annoying log-spam.

Yes, I meant the GUC_REPORT method. I don't think something like read-only and read/write is intuitive as values values of target_server_type, because people who want to use reporting apps that use temporary tables have to connect to the primary, because temp tables are not supported on standbys. I think the current notion of primary/standby is good; those who require temp tables need to specify primary.

> I'd rather not bind it directly to "in recovery" because we're likely to
> want to support read-only logical replication nodes in future, but even
> that would be OK.

Maybe we should determine the name and value of the connection parameter and GUC-REPORTed variable in light of the logical replication.

Regards
Takayuki Tsunakawa