High Load on Single TiKV Instance

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

Original topic: tikv单实例负载高

| username: magongyong

【TiDB Usage Environment】Production Environment
【TiDB Version】v5.4.3
【Reproduction Path】What operations were performed when the issue occurred
Batch concurrent reads in production business, previously normal

【Encountered Issue: Phenomenon and Impact】
Today, a single TiKV instance has high load, causing slow SQL reads from that instance, resulting in widespread slow queries.

May I ask when the TiDB automatic balancing strategy will take effect in such a situation, and whether manual intervention in the balancing strategy is required.

【Resource Configuration】Go to TiDB Dashboard - Cluster Info - Hosts and take a screenshot of this page
【Attachments: Screenshots/Logs/Monitoring】

| username: xingzhenxiang | Original post link

How many TiKV nodes are in the cluster? How are the resources allocated?

| username: magongyong | Original post link

6 servers, each with 2 instances, totaling 12 instances. All servers have the same configuration: 64 cores and 256GB of RAM.

| username: magongyong | Original post link

The single instance CPU has a maximum of 32 cores, which are fully utilized. The SQL is also reasonable, but the entire cluster’s load is concentrated on one TiKV instance.

| username: magongyong | Original post link

Would enabling follower read solve the issue?

| username: tidb菜鸟一只 | Original post link

It can be alleviated. Mainly, you need to check which table the corresponding read hotspot is on and which SQL statement it corresponds to.

| username: magongyong | Original post link

| username: zhanggame1 | Original post link

Version 5.x doesn’t have this feature, you need to upgrade.

| username: magongyong | Original post link

Okay, :ok_hand:

| username: system | Original post link

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