pgsql: Check interrupts during hot standby waits

Lists: pgsql-committerspgsql-hackers
From: Simon Riggs <simon(at)2ndQuadrant(dot)com>
To: pgsql-committers(at)postgresql(dot)org
Subject: pgsql: Check interrupts during hot standby waits
Date: 2017-01-26 19:00:34
Message-ID: [email protected]
Views: Whole Thread | Raw Message | Download mbox | Resend email
Lists: pgsql-committers pgsql-hackers

Check interrupts during hot standby waits

Branch
------
master

Details
-------
http://git.postgresql.org/pg/commitdiff/e8ee3d6b859a18d7f7375ceb9e04d256eb18aaec

Modified Files
--------------
src/backend/storage/ipc/standby.c | 2 ++
1 file changed, 2 insertions(+)


From: Andres Freund <andres(at)anarazel(dot)de>
To: Simon Riggs <simon(at)2ndQuadrant(dot)com>
Cc: pgsql-hackers(at)postgresql(dot)org
Subject: Re: [COMMITTERS] pgsql: Check interrupts during hot standby waits
Date: 2017-01-26 19:06:05
Message-ID: [email protected]
Views: Whole Thread | Raw Message | Download mbox | Resend email
Lists: pgsql-committers pgsql-hackers

Hi,

On 2017-01-26 19:00:34 +0000, Simon Riggs wrote:
> Check interrupts during hot standby waits
>
> Branch
> ------
> master
>
> Details
> -------
> http://git.postgresql.org/pg/commitdiff/e8ee3d6b859a18d7f7375ceb9e04d256eb18aaec
>
> Modified Files
> --------------
> src/backend/storage/ipc/standby.c | 2 ++
> 1 file changed, 2 insertions(+)

It seems that the actual fix here would be to replace the pg_usleep loop
with a latch wait.... Then we also don't need to needlessly loop, and
actually react promptly to signals.

Andres


From: Michael Paquier <michael(dot)paquier(at)gmail(dot)com>
To: Andres Freund <andres(at)anarazel(dot)de>
Cc: Simon Riggs <simon(at)2ndquadrant(dot)com>, PostgreSQL mailing lists <pgsql-hackers(at)postgresql(dot)org>
Subject: Re: [COMMITTERS] pgsql: Check interrupts during hot standby waits
Date: 2017-01-27 02:24:16
Message-ID: CAB7nPqTHLpuv1B71xV3KznHQjSt3Bv3Y0vpoGif1ojuVhM787A@mail.gmail.com
Views: Whole Thread | Raw Message | Download mbox | Resend email
Lists: pgsql-committers pgsql-hackers

On Fri, Jan 27, 2017 at 4:06 AM, Andres Freund <andres(at)anarazel(dot)de> wrote:
> Hi,
>
> On 2017-01-26 19:00:34 +0000, Simon Riggs wrote:
>> Check interrupts during hot standby waits
>>
>> Branch
>> ------
>> master
>>
>> Details
>> -------
>> http://git.postgresql.org/pg/commitdiff/e8ee3d6b859a18d7f7375ceb9e04d256eb18aaec
>>
>> Modified Files
>> --------------
>> src/backend/storage/ipc/standby.c | 2 ++
>> 1 file changed, 2 insertions(+)
>
> It seems that the actual fix here would be to replace the pg_usleep loop
> with a latch wait.... Then we also don't need to needlessly loop, and
> actually react promptly to signals.

For the record, a patch has been posted on the original thread:
https://www.postgresql.org/message-id/CAB7nPqRNcd+FR=8_Lyb65jBcRPX+59hXobi5G-wf4fxRak_6gw@mail.gmail.com
--
Michael