TiDB: Subsequent SQL Execution Slows Down Due to Resource Constraints on a Single Node

This topic has been translated from a Chinese forum by GPT and might contain errors.

Original topic: tidb 一个节点资源紧张引起后续sql都执行慢

| username: TiDBer_AxdzSBqt

Currently, TiKV uses region leader reads. Once a certain TiKV node is resource-constrained, submitting SQL1 to the cluster after calling the interface will query data from this resource-constrained TiKV node. The cop_task execution time will be long, indirectly causing the entire SQL query to slow down. During this period, if SQL2 is submitted to the cluster and also queries this node, it will cause SQL2 to be slow as well. How can this be avoided? Follower reads seem to only alleviate the issue but not solve it.