Re: java.net.SocketException: Connection reset by peer: socket write error

From: "Markus Wollny" <Markus(dot)Wollny(at)computec(dot)de>
To: "Oliver Jowett" <oliver(at)opencloud(dot)com>
Cc: <pgsql-jdbc(at)postgresql(dot)org>
Subject: Re: java.net.SocketException: Connection reset by peer: socket write error
Date: 2004-12-03 09:12:33
Message-ID: 2266D0630E43BB4290742247C8910575069E4580@dozer.computec.de
Views: Raw Message | Whole Thread | Download mbox | Resend email
Thread:
Lists: pgsql-jdbc

Hello!

Thank you very much for this hint - there is indeed a firewall between this in-house appserver and the DB - and that's been bugging me personally for ages, since idle PGAdminII-connects would be closed without notice, too, which causes long timeout-waits in PGAdminII and forces me to restart the connection. This might be the cause for this issue, too - and the admin of our inhouse network seems to be unable to resolve this firewall-behaviour. I'll try and see if it changes anything, if I switch off of the connection-persitancy ("Maintain connections across client requests").

Thanks again for your help!

Kind regards

Markus

> -----Ursprüngliche Nachricht-----
> Von: Oliver Jowett [mailto:oliver(at)opencloud(dot)com]
> Gesendet: Donnerstag, 2. Dezember 2004 22:26
> An: Markus Wollny
> Cc: pgsql-jdbc(at)postgresql(dot)org
> Betreff: Re: [JDBC] java.net.SocketException: Connection
> reset by peer: socket write error
>
> Markus Wollny wrote:
> > It seems
> > that somewhere along the line, the Application Server just
> looses the
> > connection to the database after about 2h17min total runtime of the
> > batch.
>
> > java.net.SocketException: Connection reset by peer: socket
> write error
> > at java.net.SocketOutputStream.socketWrite0(Native Method)
>
> Check for firewalls between the appserver and the database
> server. The error is a TCP-level connection error that means
> that the peer (the database server) unexpectedly disappeared
> on us, and we didn't see an orderly shutdown of the
> connection. If it always happens after a certain connection
> length, or after the connection has been idle for a certain
> period, it's likely that there is a stateful firewall in the
> way that is dropping the connection after that period.
>
> Failing that, does the database side log anything about the
> disconnection?
>
> -O
>

Browse pgsql-jdbc by date

  From Date Subject
Next Message Dave Smith 2004-12-03 14:13:20 Re: Bug in setFetchSize
Previous Message Iain 2004-12-03 08:09:09 Re: meta data information returned for domains