We are setting up Marvel separate monitoring cluster to monitor our ES cluster in prod.
I have some queries regarding this set up
Is there any impact of in terms of performance or resource usage in production cluster due to Marvel indexes and queries.
In case monitoring cluster is down will it make any impact on ES production cluster.
What would be the impact on cluster when we have large number of ES nodes and query from Marvel requires to aggregate data from all nodes and show it in dashboard.
We are using ES version 1.4.5 and Marvel version 1.3.1
Thanks
Virendra Agarwal
Hi there, it's difficult to verify for your particular use case, because we no longer support the versions of ES and Marvel that you're using (end-of-life info).
However, for the current version of the Elastic Stack (6.4), I can tell you that your production cluster won't be affected by your monitoring cluster going down. In terms of how the monitoring cluster will impact the performance of the production cluster, all stack products ship monitoring data through the production cluster and then into the monitoring cluster, so that small amount of work by the production cluster will utilize a small amount of resources.
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.