How to Force Offline a TiFlash Node in Pending Offline and Down Status

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

Original topic: tiflash节点处于pending offline和down状态,如何强制下线

| username: qiuxb

[TiDB Usage Environment] Production Environment / Testing / PoC
[TiDB Version]
[Reproduction Path] What operations were performed when the issue occurred
[Encountered Issue: Problem Phenomenon and Impact]
The cluster has two TiFlash nodes.
When scaling down one of the TiFlash nodes, the node status remains in a pending offline state.
The other TiFlash node automatically went down and is currently in a down state.

Currently, I have tried restarting the TiFlash node that is in the down state, but it keeps restarting after being started.
How can I forcibly take these two TiFlash nodes offline and redo them?
[Resource Configuration]
[Attachments: Screenshots/Logs/Monitoring]

| username: WalterWj | Original post link

First set the TiFlash replica to 0, and then try again.

| username: tidb菜鸟一只 | Original post link

Will TiFlash not be used in the future? Or will it be migrated? If it will not be used, you need to set all TiFlash replicas to 0 before scaling down. If it will be migrated, it is recommended to scale up first and then scale down.