TiKV Startup Error

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

Original topic: tikv启动报错

| username: TiDBer_QHSxuEa1

Excuse me, everyone. When starting TiDB, one of the TiKV nodes is in a disconnected state, and restarting still doesn’t bring up this node. How should this issue be handled?

| username: TiDBer_jYQINSnf | Original post link

This log is not from TiKV, is it? Please post the log from the TiKV node that encountered the error.

| username: xfworld | Original post link

What operations were performed on this node to make it like this?

| username: forever | Original post link

It would be best if you could share all the operations you performed before and after the cluster issue occurred, so we can better diagnose the problem.

| username: 有猫万事足 | Original post link

The error is on this line in tiup.

Run the following command to see if this metric is present:

curl http://{tikv_ip:tikv_port}/metrics | grep tikv_raftstore_region_count

| username: Jellybean | Original post link

The reason for the error in this log screenshot is the inability to evict the leader. This might just be a surface issue, and it is highly likely that the TiKV node is inaccessible, hence the failure to retrieve the relevant metric information.

The disconnected status usually occurs when the TiKV node fails to start. As for the cause, it requires a more detailed examination of the tikv.log on that node. The information provided by the original poster is too limited; it would be best to provide the tikv.log along with the relevant Grafana monitoring panels.

| username: dba远航 | Original post link

It feels like there is an issue with the physical file.