[GitHub] [hbase] bharathv commented on a change in pull request #2322: [HBASE-24956] ConnectionManager#locateRegionInMeta waits for user region lock indefinitely.

[GitBox <gi...@apache.org>]

bharathv commented on a change in pull request #2322:
URL: https://github.com/apache/hbase/pull/2322#discussion_r478687341



##########
File path: hbase-client/src/main/java/org/apache/hadoop/hbase/client/ConnectionImplementation.java
##########
@@ -968,6 +968,19 @@ private RegionLocations locateRegionInMeta(TableName tableName, byte[] row, bool
     }
   }
 
+  private void takeUserRegionLock() throws IOException {
+    try {
+      long waitTime = connectionConfig.getScannerTimeoutPeriod();
+      if (!userRegionLock.tryLock(waitTime, TimeUnit.MILLISECONDS)) {

Review comment:
       @saintstack Rushabh and I had a quick offline chat about this PR. We were wondering what is the right timeout to use for this lock. In the client code path there are a bunch of time out configurations depending on the path we take and sometimes layered on top of each other.
   
   Specifically I was wondering if hbase.client.operation.timeout would be the right one to use for this. I understand that we are using the scanner timeout here because the call wraps a scanner with the same timeout. From a client standpoint though, scanner is just an implementation detail of locateRegion (root caller in this case) and that root caller should be wrapped with a general operation timeout rather than a timeout that is specific to the scanner.
   
   Not a big deal but I was just curious and would like to know your thoughts. Thanks.




----------------------------------------------------------------
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:
users@infra.apache.org