Mutiple Queries queued due to single heavy query in Presto 0.271.1

[Sudhan T C B]

Hi All,
     I have 1 coordinator and 20 worker nodes with 247GB of memory each. we observed that a single query that takes peak memory of around 650GB due to which other queries started going to queue. can anyone help me understand why the other queries are not utilizing the memory of the remaining worker nodes?

Please help me understand how I can tackle this issue.

These are configurations applied to worker and coordinator nodes.

query.max-memory-per-node: 140GB

query.max-total-memory-per-node: 150GB 

query.max-memory: 800GB

Thanks,

Sudhan