Providing a good monitoring environment for cloud disks is an important part of maintaining high data reliability. You can use Cloud Monitor service to monitor the cloud disk mounted on CVM instance to view the metric data of cloud disk and analyze and set alarms related to cloud disk. Cloud monitor collects the raw data of hard disk from the running CVM instance and displays the data in a readable chart form. The statistical data is saved for one month by default. You can observe the situation of disks in different time periods within a month to better understand information such as usage, reading, and writing.
You can obtain data through the cloud monitor console or cloud monitor API. For more details, see Getting the Monitoring Data of Specified Metrics and Get monitoring views and reports.
Currently, Cloud Monitor provides the following monitoring metrics for CBS:
| Metric Name | Calculation Method | Meanings of Metrics | Unit | Statistical Granularity (Period) |
|---|---|---|---|---|
| Disk Read IOPS | The average value of hard disk reading IOPS in the statistical period | Frequency of IO reading from hard disk to memory per second | Frequency | 10s, 60s, and 300s |
| Disk Read Traffic | The average value of hard disk reading throughput in the statistical period | Rate of data reading from hard disk to memory | KB/s | 10s, 60s, and 300s |
| Disk Write IOPS | The average value of hard disk writing IOPS in the statistical period | Frequency of IO writing from memory to hard disk per second | Frequency | 10s, 60s, and 300s |
| Disk Write Traffic | The average value of hard disk writing throughput in the statistical period | Rate of data writing from memory to hard disk | KB/s | 10s, 60s, and 300s |
| IO Await | The average value of ioawait of hard disks in the statistical period | Percentage of time when the CPU is idle and there are still I/O requests that are not completed during the sampling period | ms | 10s, 60s, and 300s |
| IO Svctm | The average value of svctm of hard disks in the statistical period | IO service time | ms | 10s, 60s, and 300s |
| IO %util | The average value of io_util of hard disks in the statistical period | Ratio of time when the disk has IO operations (i.e., non-idle time) | % | 10s, 60s, and 300s |