Apache Pinot

I’d heard that the server process has a set (10?) number of threads to handle query requests. Assuming we have very low qps, and want to minimize latency, is there a way to set the number of threads == the number of cores? And are there any other places in the system where similar constraints exist?

I see “UnboundedResourceManager” as an alternative to “PolicyBasedResourceManager”, what would be the downside to using that one? And assuming that’s not a bad idea, how would I configure my cluster to use it?

I think it defaults to 2x the number of cores. Also, it helps improve latency by parallelizing multiple segments processing for the Amar query.

Lol, it seems like an auto correct, now I can't recall what I intended to say there

But if I understand what you’re saying, you think that all server CPU resources are currently being used for a query, so there’s no benefit to switching to UnboundedResourceManager, right?

Yeah, assuming you have enough segments, cores will be used to process them.