Hi folks, I am hitting a very strange issue when running an Elasticsearch container in an Openshift environment. Basically when the container spins up - our init scripts run - and then trigger /usr/share/elasticsearch/bin/elasticsearch at which point the Java process just stops in its track and produces no logs for exactly 6 minutes. Only log printed is:
warning: no-jdk distributions that do not bundle a JDK are deprecated and will be removed in a future release
After the 6 minutes have passed - its as if some block is lifted - and ES logs start printing as if nothing had happened.
I've tried:
- Running on different physical node.
- Giving ES unlimited resources and increasing heap.
- Testing IO performance of attached storage.
Performance tests come back normal, and none of the above alterations help in any way. Just for whatever reason - on that particular environment - there is this 6 minute timeout.
Has anyone encountered anything like this? Any pointers would be highly appreciated.