High Resource Usage Issue on a Single TiKV Node

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

Original topic: 关于tikv一个节点资源使用高的问题

| username: TiDBer_Y2d2kiJh

【TiDB Usage Environment】Production Environment
【TiDB Version】v5.4.0
【Reproduction Path】During the operation of TiDB, it was found that the resource usage of one TiKV was high. Is it because TiKV uses the Raft protocol to transfer data to other TiKVs, causing one of the TiKV servers to have high resource usage?
【Encountered Problem: Problem Phenomenon and Impact】
【Resource Configuration】Go to TiDB Dashboard - Cluster Info - Hosts and take a screenshot of this page
【Attachments: Screenshots/Logs/Monitoring】

| username: TiDBer_jYQINSnf | Original post link

If a single TiKV CPU is high, check if it’s due to reads, look at the coprocessor, and check the monitoring panel’s thread CPU to see which thread is high.

TiKV has multiple regions, which means multiraft, and each TiKV has both leaders and followers, so it’s not a situation where one TiKV is entirely passing logs to other TiKVs. It’s not like MySQL’s master-slave mode. If there are no hotspots in read and write operations, the CPU usage should be almost consistent.

| username: xfworld | Original post link

Troubleshoot hot spot issues…

Refer to:

The structure of the general data table is related, causing data to be written to only one node, and the regions are not sufficiently dispersed…

| username: tidb菜鸟一只 | Original post link

Generally, if a TiKV node has high load, it is usually a hotspot issue. Theoretically, if the data is well-balanced, the load should be evenly distributed across all nodes.

| username: system | Original post link

This topic was automatically closed 60 days after the last reply. New replies are no longer allowed.