Re: Cancel/Kill backend functions - Mailing list pgsql-patches

From Magnus Hagander
Subject Re: Cancel/Kill backend functions
Date
Msg-id 6BCB9D8A16AC4241919521715F4D8BCE34BA87@algol.sollentuna.se
Whole thread Raw
In response to Cancel/Kill backend functions  ("Magnus Hagander" <mha@sollentuna.net>)
Responses Re: Cancel/Kill backend functions  (Neil Conway <neilc@samurai.com>)
Re: Cancel/Kill backend functions  (Tom Lane <tgl@sss.pgh.pa.us>)
List pgsql-patches
>> Per previous discussions, here are two functions to send INT and TERM
>> signals to other backends.They permit only INT and TERM, and permits
>> sending only to postgresql backends (as registered in pgstat).
>
>Why does this depend on pgstat? ISTM it would be better to use the
>per-backend PGPROC information, which is stored in shared memory.
>Consider TransactionIdIsInProgress() for an example.

I guess the main reason is that I didn't find how to do it. With that
pointer, I can probably redo it.

The other thought is that you're not going to have much use of this if
you don't have pgstat anyway - how are you going to find out which
backends actually exist?

But I could certainly give it a try to recode it on that code.

//Magnus

pgsql-patches by date:

Previous
From: Manfred Koizar
Date:
Subject: Build problems
Next
From: Manfred Koizar
Date:
Subject: Tuple sampling