Note:
This topic has been translated from a Chinese forum by GPT and might contain errors.Original topic: tidb_tikvclient_backoff_seconds_count value 近万
|
username: leopardxu
[TiDB Usage Environment] Production Environment
[TiDB Version] v6.1.2
[Reproduction Path] Operations performed that led to the issue
[Encountered Issue: Issue Phenomenon and Impact]
[Resource Configuration]
[Attachments: Screenshots/Logs/Monitoring]
Can you see what the problem is? I’ve set the threshold to 3000, but it still keeps alerting. My V5.2.1 cluster doesn’t alert with a threshold of 100.
Could you share the alert information? It might be due to other tasks pending or an issue with the monitoring data.
We are using version 6.1, and it reports about ten issues a day.
Check the backoff error logs in the logs to see what the errors are.
The image you provided is not accessible. Please provide the text content you need translated.
From the provided monitoring alert screenshot, the total amount has reached the w level, but sometimes there are different types of alerts. We need to confirm whether the actual monitoring has such a high value. The monitoring screenshot you provided covers 10 hours, but the monitoring data statistics should be for more than just 10 hours; it should be for the entire period. What is the current impact? Also, looking at the bottom screenshot, it seems to be monitoring information rather than alert information. How is this strategy designed?
Did you set the alert threshold incorrectly? Shouldn’t it be the max value instead of the total value?
You need to issue the corresponding alert expression or alert information to conduct a targeted analysis.
This topic was automatically closed 60 days after the last reply. New replies are no longer allowed.
|
username: 
|
username: 
|
username: 
|
username: 
|
username: 
|
username: