Page MenuHomePhabricator

Increase HADOOP_HEAPSIZE (-Xmx) for hive-server2
Closed, ResolvedPublic1 Estimated Story Points

Description

hive-server2 on analytics1027 should run with more memory. Oliver has seen instances where persistent JDBC connections to hive-server2 can cause the JVM process to run out of memory.

This *should* be done by setting HADOOP_HEAPSIZE, however, this isn't working properly for some reason.

Event Timeline

Ottomata claimed this task.
Ottomata raised the priority of this task from to Medium.
Ottomata updated the task description. (Show Details)
Ottomata added projects: Analytics, acl*sre-team.
Ottomata changed Security from none to None.
Ottomata subscribed.
Ottomata raised the priority of this task from Medium to High.Dec 1 2014, 3:28 PM
Ottomata lowered the priority of this task from High to Medium.Dec 2 2014, 3:28 PM

Oof, we might need a new machine for hive server.

http://www.cloudera.com/content/cloudera/en/documentation/core/latest/topics/cdh_ig_hive_install.html

analytics1027 only has 8G of RAM, and is collocated with a few other services already.

Change 250013 had a related patch set uploaded (by Ottomata):
Properly export env vars from hive service default files

https://gerrit.wikimedia.org/r/250013

Change 250013 merged by Ottomata:
Properly export env vars from hive service default files

https://gerrit.wikimedia.org/r/250013

Change 250014 had a related patch set uploaded (by Ottomata):
Increase hive-server2 heap size and update cdh submodule to allow this

https://gerrit.wikimedia.org/r/250014

Change 250014 merged by Ottomata:
Increase hive-server2 heap size and update cdh submodule to allow this

https://gerrit.wikimedia.org/r/250014

Ottomata set the point value for this task to 1.
Ottomata moved this task from Next Up to Done on the Analytics-Kanban board.