Vlad Rozov created APEXCORE-589: ----------------------------------- Summary: Validate health of MiniYARNCluster in StramMiniClusterTest Key: APEXCORE-589 URL: https://issues.apache.org/jira/browse/APEXCORE-589 Project: Apache Apex Core Issue Type: Improvement Reporter: Vlad Rozov Priority: Minor
In case MiniYARNCluster is started on a node with the limited disk space (more than 90% which is the default value for yarn.nodemanager.disk-health-checker.max-disk-utilization-per-disk-percentage), it will start and be in STARTED state, but NodeManager will go into UNHEALTHY state and there will be no available NodeManagers for the Yarn mini cluster. This leads to obscure test failures with a hard to identify reason for the problem. It should be possible to check NM and RM health when Yarn mini cluster starts and fails the test. -- This message was sent by Atlassian JIRA (v6.3.4#6332)