The _health_report API provides a yellow/red status for the shard count within a cluster for the frozen and non-frozen tiers.
However, based on this, the shard count checks with the _health_report API is currently static.
This can be insufficient for large deployments that add 20-40 primary shards per hour to their clusters, requiring the users to have an alternate method to perform shard limit health checks despite the existence of an API to perform the checks.
Request
Could the check value become dynamic (percentage-based) or user-configurable via a setting?
Description
The
_health_report
API provides a yellow/red status for the shard count within a cluster for the frozen and non-frozen tiers. However, based on this, the shard count checks with the_health_report
API is currently static.This can be insufficient for large deployments that add 20-40 primary shards per hour to their clusters, requiring the users to have an alternate method to perform shard limit health checks despite the existence of an API to perform the checks.
Request
Could the check value become dynamic (percentage-based) or user-configurable via a setting?