# Cause Analysis

+ Summary:
  + 19:50 ~ 21:00 The response time of the gRPC coprocessor duration increases with read hotspots in the cluster
  + Starting at 21:00, the coprocessor Total Ops Details (Index Scan) has a clear upward trend, and it is accompanied by a hot read phenomenon
+ Root cause: A large amount of data is scanned during SQL statement execution, resulting in hot spots
  + The amount of data in the report_interact_detail table is constantly increasing
  + The SQL statement uses Order by winning_create_time DESC, and the index scan uses a single column index on the create_time field, which causes a large amount of data to be scanned during SQL execution, resulting in a hot read phenomenon.
------------------------------------------------------------
- Increased response time of gRPC coprocessor duration in the cluster from 19:50 to 21:00, accompanied by read hotspots
- Starting at 21:00, Coprocessor Total Ops Details (Index Scan) has a clear upward trend and is accompanied by a hot read phenomenon

In summary, the root cause of this problem is that a large amount of data is scanned during SQL execution, leading to hot spots

- The amount of data in the report_interact_detail table is constantly increasing
- The SQL statement uses Order by winning_create_time DESC, and the index scan uses a single column index on the create_time field, causing a large amount of data to be scanned during SQL execution, resulting in a read hot spot