# 2021-03-18-21:33:35+0800 update cluster Colopl sakura-dev-0 pd dashboard.public-path-prefix config

- Cause:
```
if cluster A restore from cluster B, cluster A pd dashboard public-path-prefix will use cluster B config.

expect Colopl sakura-dev-0 cluster pd public-path-prefix is "/proxy/cluster/1363424513567494144/pd/dashboard" but got `public-path-prefix = "/proxy/cluster/1345915126238154752/pd/dashboard"`
```

- Operation:
```bash
# take a backup of current tc
kubectl get tc -n ${NS} -o yaml > tc_backup.yaml

# edit target tidbcluster resource
kubectl edit tc -n ${NS}

public-path-prefix = "/proxy/cluster/1345915126238154752/pd/dashboard" => public-path-prefix = "/proxy/cluster/1363424513567494144/pd/dashboard"

# wait the upgrade complete
kubectl get po -n ${NS} --watch
```

- Impact:
```
The user can not view the diagnosis page
```
see detail info in https://pingcap.slack.com/archives/G017V74MQ80/p1616061377001900

- Recovery operation: 
- 
```bash
# recovery cluster config
kubectl apply tc_backup.yaml -n ${NS}
```

- Time window: TODO