Thank you @leandrojmp . Your response is very concise and accurate.
I have a related post about elastic agent datasets: Performance impact of setting 'namespace' in Elastic agent policy config
If I have 2000 Elastic agents, 20 agent policies, does that mean that:
- Look up the Elastic agents that I want to have a customized data retention policy
- For each Elastic agent, I need to look up the datastream used by that agent going to kbn:/app/fleet/data-streams
- For each datastream that I want to have a customized ILM, I would need to use custom ILM policies according to the documentation
Is my understanding correct?