d-c-manning commented on code in PR #6789:
URL: https://github.com/apache/hbase/pull/6789#discussion_r2032168923
##########
hbase-server/src/main/java/org/apache/hadoop/hbase/regionserver/HRegion.java:
##########
@@ -8438,12 +8445,12 @@ public void startRegionOperation(Operation op) throws
IOException {
throw new
NotServingRegionException(getRegionInfo().getRegionNameAsString() + " is
closing");
}
lock(lock.readLock());
- // Update regionLockHolders ONLY for any startRegionOperation call that is
invoked from
- // an RPC handler
+ // Update regionLockHolders for every startRegionOperation call that is
invoked from
+ // either an RPC handler or otherwise. We will interrupt only RPC
operations
Review Comment:
yeah I don't know the original intent - whether it was just starting with a
"safe" set of operations to interrupt, or if there is more thought beyond that.
While it is true we will abort the RS, we do the interrupt, by default, 5
times, every 10 seconds, before getting to the abort after 60 seconds.
But the question is whether these other operations are safe to interrupt. I
don't know how to answer that. It is possible for it to be "safe" to abort and
yet not "safe" to interrupt... depending on what those operations do and how
they handle interrupts.
--
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]
