virajjasani commented on a change in pull request #809: HBASE-23261 :
Processing ZK BadVersionException during node transition
URL: https://github.com/apache/hbase/pull/809#discussion_r344433462
##########
File path:
hbase-server/src/main/java/org/apache/hadoop/hbase/coordination/ZKSplitTransactionCoordination.java
##########
@@ -163,6 +166,10 @@ public void waitForSplitTransaction(final
RegionServerServices services, Region
}
Thread.sleep(100);
spins++;
+ if (spins > SPLIT_WAIT_TIMEOUT) {
Review comment:
That is correct. What used to happen so far is, even if in the first try if
setData() to ZK fails, we used to propagate BVE up the caller and then rollback
splitting, with this patch, we will retry setData() at least 10-15 times before
we throw this Exception and hence, the probability of rollback of splitting due
to BVE comes down to almost 0
----------------------------------------------------------------
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]
With regards,
Apache Git Services