Suppose we are a team with 100 other teams querying our cluster, either through Kibana or Elasticsearch APIs
Can we identify those with a special field like a uniqueID specific to one team but those fields are not indexed or queried, just pass it to every query and Elasticsearch silently ignores the field.
We will log these queries and plot based on the uniqueID.
Consider there are 100 teams hitting a single Elasticsearch cluster either through Kibana or Elasticsearch APIs. We need to identify the usage of Elasticsearch by each team.
One solution is that we could allocate one ID to one Team.
Questions:
Can we pass this ID in search queries to identify how many queries each team has hit for a day?
However we do not want this ID to be either indexed or used in scoring of each query. Can this ID be a silent field, which we can just log and calculate the frequency per Team?
The only way I can see to do this if Kibana queries need to be counted would probably be to define separate users per team and track this through audit logging.
You may also be able to do it through an external proxy but I am not aware of any way to otherwise do this directly in Elasticsearch.
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.