TiFlash directory accidentally deleted, cluster fails to start after reboot

Translated
1

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

Original topic: tiflash目录被误删，重启集群后起不来

| username: Christophe

【TiDB Usage Environment】Production\Testing Environment\POC
【TiDB Version】
【Encountered Problem】The entire directory of tiflash in one of the environments was deleted by someone, and the cluster cannot be restarted. How can this be handled? Is it enough to remove the tiflash configuration in the config?
【Reproduction Path】What operations were performed that caused the issue
【Problem Phenomenon and Impact】

2
| username: ddhe9527 | Original post link

When using tiup cluster start, use the -R option to sequentially start PD, TiKV, and TiDB. Then, scale out a new TiFlash node, and finally scale in the damaged TiFlash node.

3
| username: Christophe | Original post link

Thanks a lot! In the test environment, I directly removed TiFlash from the configuration and then restarted.

4
| username: tidb狂热爱好者 | Original post link

This method is good, and it worked successfully in the experiment.

5
| username: Christophe | Original post link

There are still some doubts. If the TiFlash node cannot start, it should not be possible to scale down, right?

6
| username: ddhe9527 | Original post link

First expand and then shrink. Even if it can’t start, you can still shrink it. After it goes DOWN, the replicachecker will start to replenish replicas on other TiFlash nodes.

7
| username: Christophe | Original post link

Thank you, I’ll give it a try when I get a chance. Previously, when the drainer went down, it couldn’t be scaled down directly using tiup.

8
| username: system | Original post link

This topic was automatically closed 1 minute after the last reply. No new replies are allowed.