I'd like to be able to find out what my users are doing, and more specifically list the users who are consuming CPU, consult the list of "big" requests and the linked user.
Basically, I'd like to know if someone is making bad requests, too big or too complex.
Is there a way to do this ? With Stack Monitoring ?
If you have access to the X-Pack features, including the Stack Monitoring feature, which are available only in the paid subscription , then most of these monitoring are available .
but if you are using OSS version then you need to build a monitoring system with combination of below alternative methods and tools :
Elasticsearch Node Stats API : Elasticsearch exposes a Node Stats API, which you can use to gather detailed statistics about each node in your cluster, including CPU usage, memory usage, indexing rates, and more. You can periodically query this API and store the data for further analysis.
Elasticsearch logs and Slow Logs : Elasticsearch provides a Slow Log feature apart from normal logging, which records queries that take longer than a specified threshold to execute. This can help you identify slow or resource-heavy queries that might be affecting the cluster's performance.
Third-Party Monitoring Tools : There are third-party monitoring tools and frameworks available that can help you monitor and analyze your Elasticsearch cluster's performance. Some popular options include Prometheus, Grafana, and Kibana (as a standalone monitoring tool).
This topic was automatically closed 28 days after the last reply. New replies are no longer allowed.