Dear list,

having recently installed Zope-2.10.2-final (I skipped some versions
before that) I stumbled across an easy way to provoke seemingly
harmless POSKeyErrors. Seemingly harmless because the copied or
imported objects seem to be ok.

The setup is quite straightforward:

    1 ZEO-Server instance with one Data.fs
    1 Zope instance mounting this DB at mount point /
    all served locally to and from the same server

Whenever I import a .zexp-file or paste a copied object, I get the
following traceback in zeo.log (reformatted for better readability).
This happens even copy-pasting something as simple as a DTML-Method.

Traceback (most recent call last):
  File "/home/zope/lib/python/ZEO/zrpc/", 
       line 518, 
       in handle_request
       ret = meth(*args)
  File "/home/zope/lib/python/ZODB/FileStorage/", 
       line 625, in modifiedInVersion
       pos = self._lookup_pos(oid)
  File "/home/zope/lib/python/ZODB/FileStorage/", 
       line514, in _lookup_pos
       raise POSKeyError(oid)
POSKeyError: 0x021e8a    

Interestingly everything works just fine when I mount the same Data.fs
directly in the Zope instance, skipping the ZEO-indirection.

Any insights, ideas or hints where to look next would be very much

Best wishes,

internetmanufaktur jo----------------------------- Berlin, Germany
  |||||||||||||||meder------------------- fon: ++49-30-44 04 27 82 -------------------- fax: ++49-30-44 04 30 95
Kollwitzstr. 66 ------------------------- mob: ++49-170-2 98 89 97
10435 Berlin --------------------------------- USt-ID# DE216313426
For more information about ZODB, see the ZODB Wiki:

ZODB-Dev mailing list  -

Reply via email to