Details
-
Improvement
-
Status: Resolved
-
Minor
-
Resolution: Fixed
-
None
-
None
Description
Currently initPauseTime is hard coded as 1000,
https://github.com/apache/hbase/blob/c853c99b209cb2dfefb8d15595d9a9a5469e24fa/hbase-server/src/main/java/org/apache/hadoop/hbase/regionserver/HRegionServer.java#L2440
But in case of ServerNotRunningYetException/PleaseHoldException/CallQueueTooBigException RegionServer may wait for a long time (In our case it was ~7 min).
Currently we calculate pause time based on the number of retries and RETRY_BACKOFF. So in a failure case RS will wait max 200 sec and wait will be always 200 sec if retries further goes beyond 13.
We should make it configurable.
Attachments
Issue Links
- links to