API calls taking more than 10s

Hi There,

We are running Elasticsearch version 8.8.1 (on prem) and have the following configuration.

3 coordinator nodes
3 dedicated master nodes
17 hot nodes
20 cold nodes
6 frozen nodes (Each frozen node has about 2000 plus shard pointers.).

API calls such as _xpack/usage (47s), _node/stats are taking longer than 10s. Has anyone experienced similar behavior ?


Scaling the frozen tier helped with the timeout issue.