Follow

Slow Analytics Run Time/AppTimelineServer crash due to OutOfMemory

If the machines in your cluster contain multiple partitions, it is possible that Ambari has configured the YARN Application Timeline Server to write to multiple locations - this will increase memory usage leading to possible OOM conditions, and also considerably negatively impact Analytics run time.

To confirm if this is the case, do the following:

  1. Go to Ambari
  2. Click on YARN
  3. Click on Configs
  4. Click on Advanced
  5. Go to Application Timeline Server
  6. Check the values of yarn.timeline-service.leveldb-state-store.path and yarn.timeline-service.leveldb-timeline-store.path to confirm that they only contain a single value each, as opposed to a comma separated list.

Note as well that this issue may be present in other areas of the overall configuration, so it may be prudent to perform a spot check to ensure that other areas are configured as expected.

Was this article helpful?
0 out of 0 found this helpful
Have more questions? Submit a request

Comments

Powered by Zendesk