How to Recover After the tidb-server File is Deleted

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

Original topic: tidb-server文件被删除后,如何恢复

| username: Hacker_lombHBWo

[TiDB Usage Environment] Production Environment
[TiDB Version] 5.3.2
[Reproduction Path] /export/tidb-deploy/tidb-4000
[Encountered Problem: Phenomenon and Impact]
The files under this path were deleted, and then restored by copying files from other identical machines. After that, the cluster was restarted using tiup cluster restart tidb-cluster -N 10.11.0.11. The port started normally.
However, the dashboard keeps showing an exception. How to recover?
[Resource Configuration]
[Attachment: Screenshot/Log/Monitoring]

| username: caiyfc | Original post link

The TiDB node doesn’t have much impact, just scale it down and then scale it up again, it’s very quick.

| username: Hacker_lombHBWo | Original post link

The current display on the dashboard is abnormal. Do we need to scale up and scale down?

| username: xingzhenxiang | Original post link

The TiDB server is a stateless node. Scale down with the force option, then scale up.

| username: caiyfc | Original post link

Scaling up or down is a relatively safe, reliable, and controllable solution. After all, this display is controlled internally by the dashboard. If you don’t want to scale up or down, you can check again after some time to see if the display is normal, and then check whether the server time of this TiDB node is aligned with the time of other machines.

| username: Hacker_lombHBWo | Original post link

Okay, thank you.

| username: Hacker_lombHBWo | Original post link

Got it, thanks.

| username: 胡杨树旁 | Original post link

The cluster is now in a normal state, but the dashboard shows an anomaly. I remember the dashboard is a feature provided by PD. Can we try restarting this PD node?

| username: TiDBer_pkQ5q1l0 | Original post link

Can tiup cluster prune clustername clean it up?

| username: 孤君888 | Original post link

Scaling down and scaling up are done, there’s no need to fear a distributed architecture.

| username: system | Original post link

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