| From: | Elvis Pranskevichus <elprans(at)gmail(dot)com> |
|---|---|
| To: | pgsql-hackers(at)postgresql(dot)org |
| Cc: | Robert Haas <robertmhaas(at)gmail(dot)com>, Peter Eisentraut <peter(dot)eisentraut(at)2ndquadrant(dot)com> |
| Subject: | Re: [PATCH v1] Add and report the new "in_hot_standby" GUC pseudo-variable. |
| Date: | 2017-03-22 21:02:57 |
| Message-ID: | 1776752.pAhDKDcDgB@hammer.magicstack.net |
| Views: | Raw Message | Whole Thread | Download mbox | Resend email |
| Thread: | |
| Lists: | pgsql-hackers |
On Wednesday, March 22, 2017 4:28:18 PM EDT Robert Haas wrote:
> On Wed, Mar 22, 2017 at 4:00 PM, Peter Eisentraut
> > I think we could use "in_recovery", which would be consistent with
> > existing naming.
>
> True.
Ironically, that was the name I originally used. I'll update the patch.
> (Jaime's question is also on point, I think.)
The main (and only) point of this patch is to avoid polling. Since
"in_recovery" is marked as GUC_REPORT, it will be sent to the client
asynchronously in a ParamStatus message. Other GUC_REPORT variables set
a good precedent.
My argument is that Hot Standby is a major mode of operation, which
significantly alters how connected clients work with the server, and
this bit of knowledge is no less important than the other GUC_REPORT
vars.
Elvis
| From | Date | Subject | |
|---|---|---|---|
| Next Message | Alvaro Herrera | 2017-03-22 21:03:45 | Re: extended statistics: n-distinct |
| Previous Message | Stephen Frost | 2017-03-22 20:43:34 | Re: PATCH: Make pg_stop_backup() archive wait optional |