huaxiangsun commented on a change in pull request #1584:
URL: https://github.com/apache/hbase/pull/1584#discussion_r418788221
##########
File path:
hbase-client/src/main/java/org/apache/hadoop/hbase/MetaTableAccessor.java
##########
@@ -1846,6 +1846,16 @@ public static void deleteMergeQualifiers(Connection
connection, final RegionInfo
qualifiers.add(qualifier);
delete.addColumns(getCatalogFamily(), qualifier,
HConstants.LATEST_TIMESTAMP);
}
+
+ // There will be race condition that a GCMultipleMergedRegionsProcedure is
scheduled while
+ // the previous GCMultipleMergedRegionsProcedure is still going on, in
this case, the second
+ // GCMultipleMergedRegionsProcedure could delete the merged region by
accident!
+ if (qualifiers.isEmpty()) {
Review comment:
@saintstack I checked the code and went through what you said above. My
understanding is that the lock prevents two procedures working on the same
region at the same time. In this specific case, it is multiple
MergeTableRegionsProcedures are submitted to clean up the merge states, so
there are already locks in place. These multiple MergeTableRegionsProcedures
are executed sequentially, because it does not check if there are mergeXXX
qualifiers to delete, the second one could delete the child region. To make
MergeTableRegionsProcedures idempotent, we need this fix here. What do you
think? There is another jira HBASE-24250 created by Andrey, where multiple
MergeTableRegionsProcedures are observed in running clusters.
----------------------------------------------------------------
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.
For queries about this service, please contact Infrastructure at:
[email protected]
