Re: (Never?) Kill Postmaster? - Mailing list pgsql-general

From Christian Schröder
Subject Re: (Never?) Kill Postmaster?
Date
Msg-id 47376AC3.7030803@deriva.de
Whole thread Raw
In response to Re: (Never?) Kill Postmaster?  (Tom Lane <tgl@sss.pgh.pa.us>)
List pgsql-general
Tom Lane wrote:
>
>> I recompiled the server with debugging symbols enabled and then did the
>> following experiment: I started a query which I knew would take some
>> time. While the query executed I disconnected my dial-up line. After
>> reconnecting the backend process was still there (still SELECTing).
>> Meanwhile the query is finished and the process is idle, but it's still
>> present.
>>
>
> That is probably not the same situation because (assuming the query
> didn't produce a lot of output) the kernel does not yet think that the
> network connection is lost irretrievably.  You'd have to wait for the
> TCP timeout interval to elapse, whereupon the kernel would report the
> connection lost (EPIPE or ECONNRESET error), whereupon we'd enter the
> code path shown above.
>

That's what I also thought, but unfortunately it doesn't seem to be the
same situation. After a while, the connection disappears and the server
log says (quite correct in my opinion):

<2007-11-11 18:45:43 CET - chschroe> LOG:  could not receive data from
client: Die Wartezeit für die Verbindung ist abgelaufen
<2007-11-11 18:45:43 CET - chschroe> LOG:  unexpected EOF on client
connection

> One thing I'm suddenly thinking might be related: didn't you mention
> that you have some process that goes around and SIGINT's backends that
> it thinks are running too long?  I'm wondering if a SIGINT event is a
> necessary component of producing the problem ...
>

Maybe. On the other hand, I sent a SIGINT to my process and it
nonetheless didn't show the strange behaviour. To test this I will
change the script so that it will leave alone the processes of one of
the users.

I think the users will create some more unkillable processes tomorrow.
As soon as I have some, I will send you the login data for the server so
that you can have a look for yourself.

Regards,
    Christian

--
Deriva GmbH                         Tel.: +49 551 489500-42
Financial IT and Consulting         Fax:  +49 551 489500-91
Hans-Böckler-Straße 2                  http://www.deriva.de
D-37079 Göttingen

Deriva CA Certificate: http://www.deriva.de/deriva-ca.cer


pgsql-general by date:

Previous
From: Tom Lane
Date:
Subject: Re: (Never?) Kill Postmaster?
Next
From: Patric Bechtel
Date:
Subject: Regression in 8.3?