Task status requests to the overlord time out frequently

Hi there,

I’m wondering if other folks have seen requests to the overlord endpoint /druid/indexer/v1/task/{task_id}/status time out frequently? I’ve been using this endpoint to submit batch ingest jobs and track the status of these jobs, but have frequently seen requests timeout.

Also what are recommended values for read and connect timeouts?

Thanks and I look forward to hearing your inputs.