sodonnel opened a new pull request, #5770:
URL: https://github.com/apache/ozone/pull/5770

   ## What changes were proposed in this pull request?
   
   The logic in ReplicationManager is increasingly complex due to various edge 
cases around unhealthy and quasi-closed containers, and the differences between 
EC and Ratis.
   
   It therefore makes sense for DatanodeAdminMonitor to call the new API 
replicationManager.checkContainerState and then check the report for 
under-replicated, rather than using the ReplicaCount objects. This means all 
the logic to decide about under-replication is confined to the replication 
manager.
   
   In making this change, I needed to make a change in the EC under-replication 
handling where the container is unhealthy / unrecoverable but also has 
decommissioning indexes, as it was not being reported as under-replicated 
previously.
   
   I also refactored the decommission monitor slightly to remove references to 
containers as "unhealthy" when they are really unclosed - renaming the metrics 
and variables to unClosed.
   
   There need to be additional testing of this change, perhaps via additional 
PRs to after this and #5746 is been committed.
   
   ## What is the link to the Apache JIRA
   
   https://issues.apache.org/jira/browse/HDDS-9902
   
   ## How was this patch tested?
   
   Some tests modified, but mostly the tests are the same as before 


-- 
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.

To unsubscribe, e-mail: [email protected]

For queries about this service, please contact Infrastructure at:
[email protected]


---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]

Reply via email to