It would be interesting for you to try this with Zope 2.10/ZODB 3.7.

In ZODB 3.7, I added an attempt to do this sort of keep-alive check. Now, clients attempt to send an empty message to the server every 30 seconds during periods of inactivity. Of course, I couldn't actually test that this fixed anything, as I didn't have a way to reproduce the problem. I didn't mention this earlier in the thread, because it didn't sound to me like this was the problem Alan was having.
Alan said he was using Zope 2.9 and this ZODB 3.6.

Jim

On Apr 18, 2007, at 10:29 AM, Paul Williams wrote:

It sounds like we have the same problem. We had contracted Tres Seaver
write us a keepalive tool to ping the server periodically.  This has
fixed our problem and we haven't had a problem in 8 days.  We used to
have this problem at least once a day.

The biggest thing is that it is seen by some as a bug in Zope or Python
since we fixed it with a keepalive.  How do we definitively clear Zeo
infrastructure? Is it somehow linked to python code not recognizing the
connection loss or is this strictly an iptables issue.  Is it a bug in
iptables or just a mis-configuration?

Thanks,
Paul Williams


-----Original Message-----
From: [EMAIL PROTECTED]
[mailto:[EMAIL PROTECTED] On Behalf
Of Alan Runyan
Sent: Sunday, April 15, 2007 11:52 AM
To: Dieter Maurer
Cc: zodb-dev@zope.org
Subject: RESOLUTION: Re: [ZODB-Dev] more lockup information /
zope2.9.6+zodb3.6.2

Thanks to Jim, Theune, Dieter and all others who weighed in on this
thread.

The problem:  ZEO Clients would lock up randomly requiring restart.

The hints: Lots of 'Connection timed out' and 'No route to host' in
ZEO Server log files.

The solution: The machine the ZEO server was running iptables.  Even
though it was
allowing the zeo server port 8100; it was filtering on all interfaces.
 I simply changed rules so port 8100 would not filter the internal
network interface - at all.  ZEO listens specifically on internal
ip/port 8100.

System has been stable for several days; without zeo server logs
containing any connection errors.  I had never had this problem before
because.. well.. we use firewalls on our customers (external
interfaces and our internal interfaces never have filtering) and this
particular sysadmin was running iptables on all public/private
interfaces; locking the machine down as much as possible.

Thanks again guys!

--
Alan Runyan
Enfold Systems, Inc.
http://www.enfoldsystems.com/
phone: +1.713.942.2377x111
fax: +1.832.201.8856
_______________________________________________
For more information about ZODB, see the ZODB Wiki:
http://www.zope.org/Wikis/ZODB/

ZODB-Dev mailing list  -  ZODB-Dev@zope.org
http://mail.zope.org/mailman/listinfo/zodb-dev
_______________________________________________
For more information about ZODB, see the ZODB Wiki:
http://www.zope.org/Wikis/ZODB/

ZODB-Dev mailing list  -  ZODB-Dev@zope.org
http://mail.zope.org/mailman/listinfo/zodb-dev

--
Jim Fulton                      mailto:[EMAIL PROTECTED]                Python 
Powered!
CTO                             (540) 361-1714                  
http://www.python.org
Zope Corporation        http://www.zope.com             http://www.zope.org



_______________________________________________
For more information about ZODB, see the ZODB Wiki:
http://www.zope.org/Wikis/ZODB/

ZODB-Dev mailing list  -  ZODB-Dev@zope.org
http://mail.zope.org/mailman/listinfo/zodb-dev

Reply via email to