@Rocky We are experiencing an intermittent issue where a specific CN node's CPU usage spikes, causing overall query latency across the cluster to increase significantly. Our current suspicion is that this may be related to a hotspot condition, where a disproportionate amount of work is being assigned to a particular CN node. Have you seen similar issues or production cases before?

References • best_practices/table_clustering.md • table_design/StarRocks_table_design.md • table_design/data_distribution/Data_distribution.md • best_practices/query_tuning/schema_tuning.md