Re: [HACKERS] logical replication and PANIC during shutdowncheckpoint in publisher - Mailing list pgsql-hackers

From Michael Paquier
Subject Re: [HACKERS] logical replication and PANIC during shutdowncheckpoint in publisher
Date
Msg-id CAB7nPqRMroFOY2eLnWOEy=hjUUdOHVrD2QY=fWSDFp1afsk-yw@mail.gmail.com
Whole thread Raw
In response to Re: [HACKERS] logical replication and PANIC during shutdowncheckpoint in publisher  (Andres Freund <andres@anarazel.de>)
List pgsql-hackers
On Tue, Jun 6, 2017 at 9:47 AM, Andres Freund <andres@anarazel.de> wrote:
> On 2017-06-05 15:30:38 +0900, Michael Paquier wrote:
>> I think that it would be interesting to be able to
>> trigger a feedback message using SIGHUP in WAL receivers, refactoring
>> at the same time SIGHUP handling for WAL receivers. It is possible for
>> example to abuse SIGHUP in autovacuum for cost parameters.
>
> Could you clarify a bit here, I can't follow?  Do you think it's
> actually a good idea to combine that with the largely mechanical patch?

Sort of. The thought here is to be able to trigger
XLogWalRcvSendReply() using a SIGHUP, even if force_reply is not
enforced. But looking again at the code, XLogWalRcvSendReply() is
processed only if data is received so sending multiple times the same
message to server would be pointless. Still, don't you think that it
would be helpful to wake up the WAL receiver at will on SIGHUP by
setting its latch? XLogWalRcvSendHSFeedback() could be triggered at
will using that.

ProcessWalRcvInterrupts() could include the checks for SIGHUP by the way...
-- 
Michael



pgsql-hackers by date:

Previous
From: Craig Ringer
Date:
Subject: Re: [HACKERS] PG10 transition tables, wCTEs and multiple operationson the same table
Next
From: Peter Eisentraut
Date:
Subject: Re: [HACKERS] Error while creating subscription when server isrunning in single user mode