feroz shaik created ATLAS-1170: ---------------------------------- Summary: JVM state determined to be unstable Key: ATLAS-1170 URL: https://issues.apache.org/jira/browse/ATLAS-1170 Project: Atlas Issue Type: Improvement Reporter: feroz shaik
Issue description: We have a 4 node cluster on AWS and after recent changes to GC strategy from CMS to G1, We have frequent shutdown of cassandra nodes throwing " JVM state determined to be unstable. Exiting forcefully due to: java.lang.OutOfMemoryError: Java heap space". The cluster has issues with the data model itself which is being worked upon. Just trying to understand why changing the GC setting not being helpful and rather causing more problem. java_heap also was changed to 16G on the nodes from earlier 8G. -- This message was sent by Atlassian JIRA (v6.3.4#6332)