LLAP daemon exception when running hive query..

[#1]

Hi,

I'm running an INSERT INTO hive query into a transactional table through beeline with execution.mode=LLAP. When I run the query twice in a row it completes successfully. But when I run it for a third time I get the following output indicating (No LLAP daemons are running):

Looking back in to ambari I see an issue reported from hive for the LLAP Application: The application reported a 'NOT RUNNING' state.

So it seems after running the simple insert query a few times, it crashes LLAP application daemon. Is this a known issue?

Only LLAP related issue in the hiveserverinteractive.log are the following lines:

2017-01-25T10:56:30,394 WARN  [HiveServer2-Handler-Pool: Thread-90]: conf.HiveConf (HiveConf.java:initialize(3814)) - HiveConf of name hive.llap.daemon.allow.permanent.fns does not exist2017-01-25T10:56:30,394 WARN  [HiveServer2-Handler-Pool: Thread-90]: conf.HiveConf (HiveConf.java:initialize(3816)) - HiveConf hive.llap.daemon.vcpus.per.instance expects INT type value

Thanks,

Mike

[#2]

just for further hive.llap.daemon.vcpus.per.instance is set to ${hive.llap.daemon.num.executors} in ambari. hive.llap.daemon.num.executors = 1.

Each node in the cluster is configured to have 4 vCores. Could this cause problems?

[#3]

The applocation logs for LLAP YARN app might have some errors indicating why it has failed.

[#4]

can you try making the configuration changes identified in this HCC post:

https://community.hortonworks.com/questions/55387/cannot-start-hiveserver2-interactive-llap.html