[
https://issues.apache.org/jira/browse/HIVE-11368?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=15237833#comment-15237833
]
Poojan Khanpara commented on HIVE-11368:
----------------------------------------
I am getting the same alerts, especially when I try to drop a big schema with
cascade. I have a six node cluster and it was working perfectly before I
upgraded latest version.
> Hive Metastore process always shows alert in Ambari UI on machines with 64
> CPU cores
> ------------------------------------------------------------------------------------
>
> Key: HIVE-11368
> URL: https://issues.apache.org/jira/browse/HIVE-11368
> Project: Hive
> Issue Type: Bug
> Components: HiveServer2
> Affects Versions: 1.2.0
> Environment: 64 CPU Core.
> Reporter: Ayappan
>
> I am running Ambari with hadoop full stack installed on a cluster setup with
> machines having 64 CPU cores. All the services are up and running. But the
> Hive Metastore process always shows alert.Checking into the alert definition
> , it says Hive command was killed due timeout after 30 seconds.
> This is below command.
> /var/lib/ambari-agent/ambari-sudo.sh su ambari-qa -l -s /bin/bash -c export
> PATH='/usr/sbin:/sbin:/usr/lib/ambari-server/*:/usr/sbin:/sbin:/usr/lib/ambari-server/*:/usr/lib64/qt-3.3/bin:/usr/local/sbin:/usr/local/bin:/sbin:/bin:/usr/sbin:/usr/bin:/root/bin:/var/lib/ambari-agent:/var/lib/ambari-agent:/bin/:/usr/bin/:/usr/sbin/:/usr/iop/current/hive-metastore/bin'
> ; ulimit -s 10240 ; export
> HIVE_CONF_DIR='/usr/iop/current/hive-metastore/conf/conf.server' ; hive
> --hiveconf hive.metastore.uris=thrift://birhel17.rtp.raleigh.ibm.com:9083
> --hiveconf hive.metastore.client.connect.retry.delay=1s
> --hiveconf hive.metastore.failure.retries=1 --hiveconf
> hive.metastore.connect.retries=1 --hiveconf
> hive.metastore.client.socket.timeout=14s --hiveconf
> hive.execution.engine=mr -e 'show databases;'
> And the alert-metastore python script has a timeout of 30 seconds but the
> above Hive command takes more than 30 seconds on a 64 core machine. So it
> always shows the alert.
> Even manually running the command from command line takes lot of time (around
> 27 secs) in 64 core compared to 8 core machine (takes only 3 secs)
> Do we need to change some hive parameters ( like worker.threads ) for 64 core
> machines ?
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
