ERROR(200) KeywordIndex unindex_object could not remove documentId -702398584 from index workitems_from. This should not happen.
We are seeing traceback in the event log like this: 2005-09-09T09:28:32 ERROR(200) KeywordIndex unindex_object could not remove documentId -702398584 from index w orkitems_from. This should not happen. Traceback (most recent call last): File "/home/zope/zope-2.7.4/lib/python/Products/PluginIndexes/common/UnIndex.py", line 185, in removeForward IndexEntry indexRow.remove(documentId) File "/home/zope/zope-2.7.4/lib/python/ZODB/Connection.py", line 599, in setstate invalid = self._is_invalidated(obj) File "/home/zope/zope-2.7.4/lib/python/ZODB/Connection.py", line 628, in _is_invalidated raise ReadConflictError(object=obj) ReadConflictError: database read conflict error (oid 0x38f00c, class BTrees.IIBTree.IISet) What is the meaning of the first line? Is it something we should worry about? (workitems_from is an index of the ZCatalog, of course). -- Icube Srl http://www.icube.it/
This is actually nothing to worry about. The error message doesn't take into account that the error it's reporting could be a conflict error (these are normal errors seen during Zope execution). The catalog probably shouldn't bother logging the error in this case. On Sep 9, 2005, at 4:16 AM, Marco Bizzarri wrote:
We are seeing traceback in the event log like this:
2005-09-09T09:28:32 ERROR(200) KeywordIndex unindex_object could not remove documentId -702398584 from index w orkitems_from. This should not happen. Traceback (most recent call last): File "/home/zope/zope-2.7.4/lib/python/Products/PluginIndexes/ common/UnIndex.py", line 185, in removeForward IndexEntry indexRow.remove(documentId) File "/home/zope/zope-2.7.4/lib/python/ZODB/Connection.py", line 599, in setstate invalid = self._is_invalidated(obj) File "/home/zope/zope-2.7.4/lib/python/ZODB/Connection.py", line 628, in _is_invalidated raise ReadConflictError(object=obj) ReadConflictError: database read conflict error (oid 0x38f00c, class BTrees.IIBTree.IISet)
What is the meaning of the first line? Is it something we should worry about? (workitems_from is an index of the ZCatalog, of course). -- Icube Srl http://www.icube.it/ _______________________________________________ Zope maillist - Zope@zope.org http://mail.zope.org/mailman/listinfo/zope ** No cross posts or HTML encoding! ** (Related lists - http://mail.zope.org/mailman/listinfo/zope-announce http://mail.zope.org/mailman/listinfo/zope-dev )
Chris McDonough wrote:
This is actually nothing to worry about. The error message doesn't take into account that the error it's reporting could be a conflict error (these are normal errors seen during Zope execution). The catalog probably shouldn't bother logging the error in this case.
Well, I think this has been fixed, but it's not so benign ;-) By catching a ConflictError like this, I'm pretty sure you can end up with inconsistent index state saved in your ZODB :-( Chris -- Simplistix - Content Management, Zope & Python Consulting - http://www.simplistix.co.uk
-----BEGIN PGP SIGNED MESSAGE----- Hash: SHA1 Chris Withers wrote:
Chris McDonough wrote:
This is actually nothing to worry about. The error message doesn't take into account that the error it's reporting could be a conflict error (these are normal errors seen during Zope execution). The catalog probably shouldn't bother logging the error in this case.
Well, I think this has been fixed, but it's not so benign ;-) By catching a ConflictError like this, I'm pretty sure you can end up with inconsistent index state saved in your ZODB :-(
That method now has an explicit 'except ConflictError: raise' a few lines above the point which does the logging; I checked in that fix on 2005/07/05 (collector #1832). That fix was included in Zope 2.7.7 and in Zope 2.8.1. I would recommend that Marco upgrade to avoid the kind of consistency issues you point out. Tres. - -- =================================================================== Tres Seaver +1 202-558-7113 tseaver@palladion.com Palladion Software "Excellence by Design" http://palladion.com -----BEGIN PGP SIGNATURE----- Version: GnuPG v1.2.5 (GNU/Linux) Comment: Using GnuPG with Thunderbird - http://enigmail.mozdev.org iD8DBQFDJXdR+gerLs4ltQ4RAme2AJ979YepexVREmLWKkC1TvQK3ib2rwCeOSlE X2S0SDfij1Xx+g4x3vZ5w+Q= =IyX7 -----END PGP SIGNATURE-----
Marco Bizzarri wrote:
We are seeing traceback in the event log like this:
2005-09-09T09:28:32 ERROR(200) KeywordIndex unindex_object could not remove documentId -702398584 from index w orkitems_from. This should not happen. Traceback (most recent call last): File "/home/zope/zope-2.7.4/lib/python/Products/PluginIndexes/common/UnIndex.py", line 185, in removeForward IndexEntry indexRow.remove(documentId) File "/home/zope/zope-2.7.4/lib/python/ZODB/Connection.py", line 599, in setstate invalid = self._is_invalidated(obj) File "/home/zope/zope-2.7.4/lib/python/ZODB/Connection.py", line 628, in _is_invalidated raise ReadConflictError(object=obj) ReadConflictError: database read conflict error (oid 0x38f00c, class BTrees.IIBTree.IISet)
What is the meaning of the first line? Is it something we should worry about? (workitems_from is an index of the ZCatalog, of course).
This spurious error message has been fixed in Zope 2.7.7 last July (http://www.zope.org/Collectors/Zope/1832). Florent -- Florent Guillaume, Nuxeo (Paris, France) CTO, Director of R&D +33 1 40 33 71 59 http://nuxeo.com fg@nuxeo.com
Marco Bizzarri wrote at 2005-9-9 10:16 +0200:
We are seeing traceback in the event log like this:
There are more bugs in "KeywordIndex" (as someone else noted, this one has been fixed recently). You may consider my "ManagableIndex" and its "KeywordIndex". It has less bugs than the Zope one -- and should you find one, the fix will almost surely be much quicker than for Zope indexes ;-) <http://www.dieter.handshake.de/pyprojects/zope> -- Dieter
participants (6)
-
Chris McDonough -
Chris Withers -
Dieter Maurer -
Florent Guillaume -
Marco Bizzarri -
Tres Seaver