No alive nodes found in cluster

Sulcalibur · April 23, 2019, 11:51am

I have installed Elasticsearch on my Digital Ocean Plesk server to work with Active Collab. It all seems fine apart from after a few seconds it seems to stop.

Trying to keep the Elasticsearch nodes alive but it keeps stopping, every damn time. Any advice on this

I try: sudo service elasticsearch start
then: curl -X GET "localhost:9200/"
I get a response, but only briefly, then it stops and I get the message:
no alive nodes found in cluster in Active Collab

DavidTurner · April 23, 2019, 12:23pm

There should be some information about what's going on in the Elasticsearch server logs. Can you share them here?

Sulcalibur · April 23, 2019, 12:47pm

Hi @DavidTurner.

Thanks for the help. I have a few logs listed in the directory, is there any particular ones you need?

DavidTurner · April 23, 2019, 12:54pm

Yes, the contents of elasticsearch.log would be very helpful. I think they'll be too large to post inline but can you share them as a Gist?

Sulcalibur · April 23, 2019, 1:02pm

@DavidTurner - Here you go buddy - https://gist.github.com/Sulcalibur/76b67ff75d113f5e866de53d3482218f

DavidTurner · April 23, 2019, 1:33pm

Thanks, that looks kinda normal except it's much slower than usual, and I wouldn't expect GC warnings in a cluster with no indices. What are the specs of the host machine (RAM, CPU, etc)? Also it doesn't look like the node shut down cleanly. Does dmesg indicate it was terminated by the kernel (e.g. the OOM killer)?

Have you set the heap size to less than 50% of the total RAM? Have you disabled swap?

system · May 21, 2019, 1:33pm

This topic was automatically closed 28 days after the last reply. New replies are no longer allowed.