However, I do not see the updated settings when I get the cluster settings. How can I retrieve these settings, so I know it has correctly been updated?
I just checked this on our upcoming 5.0 release which has reworked the way settings updates are validated and trying to update that setting there results in an error message saying that [discovery.zen.fd.ping_timeout] is not dynamically updateable. We don't do this kind of settings validation on 1.x and 2.x yet so if you try to update a non updateable setting you unfortunately don't get an error. Its just silently ignored. If you don't see the updated settings in the response chances are they have not been applied.
When trying this on any 2.x version of ES you should get a warning in the logs like [WARN ][action.admin.cluster.settings] [Anything] ignoring persistent setting [discovery.zen.fd.ping_timeout], not dynamically updateable
The reason is because our NAS shared storage solution (yes, I am also banging my head against the wall) sometimes gives timeouts after more than 30 seconds, and I do not want nodes to leave the cluster yet, but give it more time (a few minutes), so existing processes like indexing do not immediately quit.
Apache, Apache Lucene, Apache Hadoop, Hadoop, HDFS and the yellow elephant
logo are trademarks of the
Apache Software Foundation
in the United States and/or other countries.