Bug #3038
Some resource map documents not getting copied to each CN
100%
Description
Issue is for tracking resource maps on CN production nodes. It appears that some set of resource map documents are not getting 'replicated' by metacat to each CN.
This is discovered by inspecting the index-processor log file - which logs messages when a file path is present in the hzCast cluster data structure but unavailable on the actual file system. During initial replication there will be gaps between when a document is added to hzCast before it is present on each CN. This issue is not concerned with that gap - this issue is concerned when a long period of time passes and the documents are still not available.
Attaching log files from this morning that indicate which resource map documents are not present by CN
History
#1 Updated by Ben Leinfelder over 12 years ago
I'm not sure if Chris or Robert ever invoked the "Get All" action from each of the three CN's replication panel. This should ensure any missing OREs (data files) are replicated across the three CNs. Before investigating this further, I'd make sure this is run first.
#2 Updated by Ben Leinfelder over 12 years ago
- Target version set to Sprint-2012.27-Block.4.2
#3 Updated by Dave Vieglais over 12 years ago
- Position set to 3
#4 Updated by Skye Roseboom over 12 years ago
Scanned cn production logs on A.M. of 7-9 and all index logs are clear of errors. No issues regarding missing ORE docs.
Will likely want to re-check this issue when the next round of ORE docs are added to production.
#5 Updated by Ben Leinfelder over 12 years ago
- Status changed from New to Closed
Robert manually forced replication on this last batch because of timeouts for Metacat replication processing.
