Severe Performance Degradation After TiKV Scaling Up or Down!

Translated
1

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

Original topic: tikv 扩容缩容后性能下降严重！

| username: TiDBer_QNwWQtiS

To improve efficiency, please provide the following information. A clear problem description can help resolve the issue faster:

[Overview] Severe performance degradation after TiKV scaling up or down

[Application Framework and Business Logic Adaptation] Java application consuming MQ data and writing to TiDB database

[Background]

[Phenomenon] Performance was acceptable before, but it became extremely poor after scaling up or down!

[Problem] The current issue encountered

[Business Impact]

[TiDB Version] 5.1

[Attachments] Relevant logs and monitoring

2
| username: tidb菜鸟一只 | Original post link

Check if it is a hot issue.

3
| username: Billmay表妹 | Original post link

The information provided is a bit limited. Check some metrics on the monitoring page~

How many TiKV nodes do you currently have?

4
| username: Billmay表妹 | Original post link

Check if there are concurrent writes and queries during peak periods, and see if there are any empty regions in the monitoring. You can try removing the scheduling of hotspot regions.

5
| username: zhanggame1 | Original post link

Let’s see the TiKV status with a Grafana screenshot.

6
| username: Anna | Original post link
  • TiDB-Overview Grafana monitoring has System info at the bottom.