| From: | Michael Paquier <michael(at)paquier(dot)xyz> |
|---|---|
| To: | Haribabu Kommi <kommi(dot)haribabu(at)gmail(dot)com> |
| Cc: | laurenz(dot)albe(at)cybertec(dot)at, Tom Lane <tgl(at)sss(dot)pgh(dot)pa(dot)us>, Robert Haas <robertmhaas(at)gmail(dot)com>, Jing Wang <jingwangian(at)gmail(dot)com>, Pg Hackers <pgsql-hackers(at)postgresql(dot)org>, "Tsunakawa, Takayuki" <tsunakawa(dot)takay(at)jp(dot)fujitsu(dot)com> |
| Subject: | Re: Libpq support to connect to standby server as priority |
| Date: | 2018-11-17 00:14:09 |
| Message-ID: | 20181117001409.GB1182@paquier.xyz |
| Views: | Raw Message | Whole Thread | Download mbox | Resend email |
| Thread: | |
| Lists: | pgsql-hackers |
On Sat, Nov 17, 2018 at 10:41:54AM +1100, Haribabu Kommi wrote:
> Yes, we need either session open or reconnect it approach to find out
> the whether server is read-write or read-only.
Even if there is no agreement on this part, wouldn't a read-only option
be enough to support any case? With a cluster of two nodes, one primary
and one standby, a connection string listing both nodes would fail only
in the middle of a planned failover. If you rinse and repeat it is
possible to have a larger control at application level because you
precisely know the whole state of the cluster at a given instant.
> And also for read-only or prefer-read connection types, once after the
> connection establishment is done, later server promotes to read-write,
> I feel we can continue the connection, that decision makes the feature
> simple, or do we want to stop the connection?
That feels like something the application needs to care about once the
session is established, but that's only my take on the matter.
--
Michael
| From | Date | Subject | |
|---|---|---|---|
| Next Message | Andres Freund | 2018-11-17 00:45:25 | Re: Why do pg_upgrade's test use the serial schedule? (actual thread) |
| Previous Message | Michael Paquier | 2018-11-17 00:06:43 | Re: ATTACH/DETACH PARTITION CONCURRENTLY |