Case studies

This section offers some real-world problem scenarios and solutions to use Elasticsearch.

Node configuration

You have a five-node production cluster, where each node has 32GB of total memory and 16GB is allocated to Elasticsearch. Lately, you've noticed a problem: every couple of days, node-05 leaves the cluster without warning. Restarting Elasticsearch on this node solves the problem temporarily, but the node will drop out of the cluster again in a few days. How do we go about looking into this issue?

The next time this error happens, check the Elasticsearch logs before restarting the node:

tail -n 500  /var/log/elasticsearch/*.log

You notice in the log file that Elasticsearch is throwing an OutOfMemoryError exception, like the following:

Caused by: java.lang.OutOfMemoryError: Java heap space
        at org.apache.lucene.store.DataOutput.copyBytes(DataOutput.java:273)
        at org.apache.lucene.util.fst.FST.<init>(FST.java:342)
        at org.apache.lucene.util.fst.FST.&lt...