How to Clear TiKV Cache After Incorrect TiKV Decommissioning Steps

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

Original topic: tikv下线步骤错误怎么清理集群下线tikv缓存

| username: 普罗米修斯

[TiDB Usage Environment] Production Environment
[TiDB Version] v5.2.4
[Reproduction Path] A TiKV node was taken offline, its status has changed to tombstone, but tiup cluster prune xx has not been executed yet. I executed store remove-tombstone in pd-ctl, and now when I go back to execute tiup cluster prune xx, it prompts “Error: no store matching address ‘192.168.80.212:20161’ found”. How can I manually clear the offline TiKV from the cluster cache display?
[Attachments: Screenshots/Logs/Monitoring]


| username: WalterWj | Original post link

Prune should be performing the manual operations you mentioned…
You can go to the .tiup directory, find the meta file for your cluster, and manually delete the information for this tikv. Not sure if there are any other risks.

| username: 普罗米修斯 | Original post link

Prune is a manual operation and should be done before store remove-tombstone. The current execution order is incorrect, and when prune is executed again, the node cannot be found.

| username: oceanzhang | Original post link

How about trying to add force?

| username: dba远航 | Original post link

Display it to see if the current situation is normal.

| username: 普罗米修斯 | Original post link

Normally, it’s above.

| username: 小龙虾爱大龙虾 | Original post link

It is indeed confirmed that this node no longer exists, you can try tiup cluster scale-in xxx --force.

| username: tidb菜鸟一只 | Original post link

Directly use --fore to clean up the meta files.

| username: 普罗米修斯 | Original post link

Yes, you can use scale-in --force.

| username: 普罗米修斯 | Original post link

No problem with trying this scale-in force.

| username: system | Original post link

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