Details
-
Bug
-
Resolution: Unresolved
-
Minor
-
None
-
None
-
None
Description
Sometimes the scheduling of a specific selective harvest is so, that one harvest is not finished before the next one is scheduled.
And so a deduplication-index is requested partially based on an unfinished harvest X.
This indexer then fails partially to create index, because it cannot retrieve metadata for harvest X.
We then see warnings like this in IndexApplication log:
Feb 14, 2012 2:44:42 PM dk.netarkivet.archive.indexserver.RawMetadataCache cacheData INFO: No data found for job '143450' for 'crawllog' in local bitarchive 'KB'. Trying other replicas. Feb 14, 2012 2:44:43 PM dk.netarkivet.archive.indexserver.RawMetadataCache cacheData WARNING: No data found for job '143450' for 'crawllog' in any replica. Feb 14, 2012 2:44:43 PM dk.netarkivet.archive.indexserver.distribute.IndexRequestServer doGenerateIndex WARNING: Failed generating index of type 'DEDUP_CRAWL_LOG' for the jobs [143449,143450]. Missing data for jobs [143450].