abstractdog commented on code in PR #5649:
URL: https://github.com/apache/hive/pull/5649#discussion_r1993615478
##########
ql/src/java/org/apache/hadoop/hive/ql/exec/tez/WorkloadManager.java:
##########
@@ -486,15 +486,16 @@ private void scheduleWork(WmThreadSyncWork context) {
LOG.info("Invoking KillQuery for " + queryId + ": " + reason);
try {
UserGroupInformation ugi = UserGroupInformation.getCurrentUser();
- SessionState ss = new SessionState(new HiveConf(),
ugi.getShortUserName());
- ss.setIsHiveServerQuery(true);
- SessionState.start(ss);
- kq.killQuery(queryId, reason, toKill.getConf());
- addKillQueryResult(toKill, true);
- killCtx.killSessionFuture.set(true);
- wmEvent.endEvent(toKill);
- LOG.debug("Killed " + queryId);
- return;
+ try (SessionState ss = new SessionState(new HiveConf(),
ugi.getShortUserName())) {
Review Comment:
here, as I referred to in the case of the MV registry, what about trying a
SessionState field in the workload manager, so instead of:
```
UserGroupInformation ugi =
UserGroupInformation.getCurrentUser();
SessionState ss = new SessionState(new HiveConf(),
ugi.getShortUserName());
```
we can initialize the SessionState when the WorkloadManager is created,
because neither the object returned by new HiveConf(), nor the
UserGroupInformation.getCurrentUser() is supposed to change over time in HS2,
it's safe to reuse SessionState
--
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: gitbox-unsubscr...@hive.apache.org
For queries about this service, please contact Infrastructure at:
us...@infra.apache.org
---------------------------------------------------------------------
To unsubscribe, e-mail: gitbox-unsubscr...@hive.apache.org
For additional commands, e-mail: gitbox-h...@hive.apache.org
