I just got marvel up and running and everything looks good except the nodes and indices keep getting grayed out. A little ! symbol is next to the nodes and indices saying that it hasnt received an update in over 1 minute. Due to the size of the cluster marvel is automatically adjusting the refresh rate to 2 minutes. It wont allow me to change it. How can I either:
Increase the timeout to over 2 minutes on the alert to keep the nodes/indices from being grayed out.
Or
Force marvel/kibana to accept a refresh rate on the dashboard of 1 minute?
10 nodes with 3.84TB of data. I have seen one or two posts where this warning pops up with small clusters as well. Unfortunately, they did not post the cause or what was done to resolve it.
I purchased 10 license for a 10 node cluster. We have a separate 4 node cluster we use for logging that I am sending all of the data to. I logged into marvel on the logging cluster and applied the license there. Could this be a licensing issue? Do the nodes in the logging cluster count towards the licensing for marvel?
I did see that time drift was the main cause. All servers are running NTP and the time is synced.
Alright so if licensing isn't and issue and all of the times on the servers are in sync, what can I do to fix this? Is it possible to force the refresh rate to 1 minute? Or is it possible to change the time that marvel alerts on the servers to 2 minutes?
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.