Ever since I upgrade my clusters to ES5.2, I have been plagued with random crashes of Elasticsearch. The only thing I see in the logs is this entry in the messages log, there is nothing in the elasticsearch log at the time of the crash:
elasticsearch.service: main process exited, code=killed, status=6/ABRT
I search on here and found some old threads talking about this, but there didn't appear to be any clear root cause. Some said it went away of a hardware change, others after upgrading the JVM. These nodes have been running ES2.2 for over a year with an issue and they are all running JVM 1.8.77. I haven't upgraded the JVM. Is there a bug in this JVM that has only been exposed by ES5 perhaps? I'm scratching my head a little trying to figure out how to stabilise the cluster.