Exception
Log in to the TDSQL console and click an instance ID to enter the instance management page. On the Monitoring and Alarms tab, you can view three IO metrics. |
| | Maximum IO utilization of the source node |
| | |
| | |
In actual use, the above three IO metrics may increase abnormally even during off-peak hours. Alarms configured for them, if any, will be triggered.
Exception causes
Possible causes of the abnormal IO increase include:
There is a calculation problem with the IO metrics, and the actually displayed values are IO values of the server where the instance and node reside.
The IO levels of other instances on the same server increase, which may affect the IO monitoring of the current instance.
Solution
This issue involves alarms. We have located it and are fixing it. You can try the following workarounds for the time being: Do not configure alarms for these three IO metrics. Or, increase their alarm thresholds to avoid the impact of abnormal alarms.
Ignore these three IO metrics when they increase abnormally and don't match the business usage.
We are monitoring and handling the IO usage issue uniformly to ensure the normal operations of instances.