Metacat

Hi, Mike:

I observed some error message during the replication from LTER to KNB:
Failed to write doc pisco_recruitment.314.1 into db because User does not have permission to update of access rules for data file pisco_recruitment.120 [edu.ucsb.nceas.metacat.ReplicationHandler]
Metacat: [ERROR]: error to handle update doc in xml_documents in time replicationUser does not have permission to update of access rules for data file pisco_recruitment.120 [edu.ucsb.nceas.metacat.ReplicationHandler]
Metacat: [ERROR]: Error in Eml200SAXHanlder.handleOnlineUrlDataFile is User does not have permission to update of access rules for data file pisco_recruitment.120 [edu.ucsb.nceas.metacat.Eml200SAXHandler]

I think this is an issue ( which may NOT relative to Margaret's one). During the replication, there no user login, so the user is null and it should have all permission to write documents into the db. Obviously, this access control rule was modified.

checked on replication for the 6 packages listed when this bug was opened. Replication for these is at 50%. 3 out of 5 eml docs have now replicated from LTER to KNB. These 2 have not:
sbc_eml210_test.4
sbc_eml210_test.5

and this package has not yet replicated from KNB to LTER:
sbc_eml210_test.6

Did you see any access control error which I reported on the second comment during the replication from lter to knb?

Moved the replication data permissions error to a separate ticket #3908

sbc_eml210_test.4 and sbc_eml210_test.5 are definitely erroring out on the knb side with an error described in ticket 3906

This was related to but 3908, where there were permissions issues with the replication handler. This caused a chain of events.

1) the replicting server wound up with zero length files on disk
2) subsequent replications tried to read those zero length files and failed

Created a replication user which has ALL permissions. Also added code to make sure that zero length files do not get written or read.

Original Bugzilla ID was 3889