Details
-
Bug
-
Resolution: Fixed
-
Minor
-
5.1, 5.2.2
-
None
Description
When an error occurs during metadatafile generation, the whole job is moved to the oldjobs folder, and a new job is started subsequently - also when a shutdown.txt is present in the main folder of the harvestercontroller.
This gives us the risk of uploading incomplete or corrupt metadata warcfiles when executing the upload_failed_warcs.sh script
Instead it should close down the harvester and send one or more notifications: One that errors were encountered durign metadatageneration, two: That the harvester were shutdown at this time and date