Note:
This topic has been translated from a Chinese forum by GPT and might contain errors.Original topic: pd节点一直down机
|
username: whale
[TiDB Usage Environment] Testing/
[TiDB Version] v5.4
[Reproduction Path] Operations performed that led to the issue
[Encountered Issue: Issue Phenomenon and Impact]
PD node alternates between up and down states
[Resource Configuration]
[Attachments: Screenshots/Logs/Monitoring]
I think the problem is that the tidb-server process is not running. You can check if the tidb-server process is running by using the ps command. If it is not running, you can start it using the systemctl start tidb command.
The image you provided is not accessible. Please provide the text content you need translated.
At least three nodes are required, currently unable to elect.
I only have a single-node PD? Just one, because it’s a test environment. It was working fine before, why does a single node need an election?
Judging by the screenshot, it seems there might be an issue with the status of TiKV. How about checking the TiKV logs?
It is recommended to provide the PD logs. There must be error logs when PD goes down or up. Check the relevant error logs. Just looking at the warning logs in TiDB is not very valuable.
A single node is fine as long as there are no issues. It seems that your node has disk I/O problems. For mixed deployments, it is recommended to check the I/O.
Your issue has been confirmed to be due to insufficient disk I/O.
This topic was automatically closed 60 days after the last reply. New replies are no longer allowed.
|
username: 
|
username: 
|
username: 
|
username: 
|
username: