I redirected the response to "zope@zope.org" where is it better placed. Gerhard Schmidt wrote at 2004-1-14 16:19 +0100:

we are running an zope system with one ZEO and six Zope Frontends

The system ist running for one Month without a problem under zope2.7b3

Today I had several errors in the zeo.log.

2004-01-14T16:01:34 ERROR(200) Conflict Resolution Unexpected error Traceback (most recent call last): File "/usr/local/www/Zope/lib/python/ZODB/ConflictResolution.py", line 109, in tryT oResolveConflict old = state(self, oid, oldSerial, prfactory) File "/usr/local/www/Zope/lib/python/ZODB/ConflictResolution.py", line 38, in state p = p or self.loadSerial(oid, serial) File "/usr/local/www/Zope/lib/python/ZODB/FileStorage.py", line 712, in loadSerial raise POSKeyError(serial) POSKeyError: 0000000000000000

This can happen when you recently packed the ZODB. In this case, the original version of the object may no longer be available. Beside the log entry above, this should have no other effect.

Soon all fronends run out of service and this error is logged.

Do you see any log entries in the frontends log files? The exception in the server should cause a "ConflictError" in one of its clients. This client should retry its request. All other clients should be unaffected. Or, to say it differently, your frontends should not die due to the above problem.

I have rebuild the Data.fs with fsrecover.py. No Errors reported but its runnig again. But still some of the above errors are logged in zeo.log.

Anybody some Idea what caused this failure.

Do you pack your database continuously? Usually (infrequent packing), one would expect that the problem above would occur with a very small probability... -- Dieter