Note:
This topic has been translated from a Chinese forum by GPT and might contain errors.Original topic: tidb日志猛增
|
username: TiDBer_2lwED25q
[Issue Encountered]: In the testing environment, a colleague deployed TiDB, 3 TiKV, and Prometheus on a virtual machine with a 200G disk. Starting this Tuesday, disk space warnings began to appear, and it was found that the TiDB log files were too large, reaching 70G. After manually deleting and retaining only the logs from the last three days, there were still about 40G left. Now, even with only the last two days’ logs retained, there are still 30G. This situation did not occur before, and the logs have increased sharply. Upon entering the log directory of the TiDB component, it was found that a log file is generated every ten minutes or so, with each log file being 300M. As of June 27, there are already nearly 100 log files.
|
username: 
|
username: 
|
username: 
|
username: 
|
username: 
|
username: 
|
username: 
|
username: 
|
username: 
|
username: 
|
username: 
|
username: 
|
username: 
If the hardware configuration allows, it is recommended not to set it to error, as it can be quite troublesome when troubleshooting issues. Some SQL statements that report errors are logged as warnings. I have encountered situations where SQL execution errors were reported in the monitoring, but they were not visible in the logs. After researching, I found that such SQL-level errors are logged as warnings.