After PD Rebuild, Database Cannot Be Accessed Normally

Translated
1

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

Original topic: pd重建后数据库无法正常访问

| username: MagicJie

[TiDB Usage Environment] Test Environment
[TiDB Version] V6.5 deployed via operator
[Reproduction Path]
Changed TiDBCluster configuration (added QoS settings, no other changes)
After deleting the related pod, PD could not start normally. Attempted recovery according to:

[Encountered Issue: Symptoms and Impact]
Pod status is normal

image
image1380×430 36.5 KB

However, the database cannot be accessed normally.

Cluster configuration files and logs are attached
[Resource Configuration]
pre-test (1).yaml (3.3 KB)

[Attachments: Screenshots/Logs/Monitoring]

image
image1380×170 75.4 KB

pre-test-tidb-0_tidb.log (5.1 MB)
pre-test-pd-0_pd (1).log (502.7 KB)
pre-test-tikv-0_tikv (1).log (5.5 MB)

2
| username: MagicJie | Original post link

I have tried deleting the cluster and redeploying it, but the same issue persists.

3
| username: yiduoyunQ | Original post link

If it is a test environment, it is recommended to delete the PV and the corresponding data and then rebuild the cluster.

4
| username: MagicJie | Original post link

There are many test cases used by the testing department, and it would be quite troublesome if the data is lost. Currently, the data in TiKV is still there, let’s see if it can be salvaged.

5
| username: yiduoyunQ | Original post link

PD disaster recovery refers to the official documentation 使用 PD Recover 恢复 PD 集群 | PingCAP 文档中心, the operation is mature.

6
| username: MagicJie | Original post link

This operation has already been tried, and now all the Pods have started, but TiDB is reporting an error when querying data.