A good monitoring environment is the key to ensuring high availability of cloud disks. You can use Cloud Monitor (CM) to monitor cloud disks attached to CVM instances, view cloud disk metrics, and analyze and configure cloud disk alarms. CM collects raw data of the cloud disk from running CVM instances and displays data through easy-to-read charts. By default, the data is retained for one month, during which you can view the disk conditions in different time periods to better know its usage and read/write operations.
You can get data in the CM console or via API. For more information, see Getting the Monitoring Data of Specified Metrics and Getting Monitoring View and Chart.
Currently, CM provides the following metrics to monitor cloud disks:
Metric | Name | Calculation | Description | Unit | Statistical Granularity (Period) |
---|---|---|---|---|---|
DiskReadIops | Disk Read IOPS | Average of read IOPS of the disk during the statistical period | Number of I/O reads from the disk to the memory every second | Count | 10s, 60s, 300s |
DiskReadTraffic | Disk Read Traffic | Average of read throughput of the disk during the statistical period | Speed of data reads from the disk to the memory | KB/s | 10s, 60s, 300s |
DiskWriteIops | Disk Write IOPS | Average of write IOPS of the disk during the statistical period | Number of I/O writes from the memory to the disk every second | Count | 10s, 60s, 300s |
DiskWriteTraffic | Disk Write Traffic | Average of write throughput of the disk during the statistical period | Speed of data writes from the memory to the disk | KB/s | 10s, 60s, 300s |
DiskAwait | Disk I/O Wait Time | Average of I/O wait time of the disk during the statistical period | Percentage of time during which the CPU is free and there are unfinished I/O requests during the statistical period | ms | 10s, 60s, 300s |
DiskSvctm | Disk I/O Service Time | Average of service time of the disk during the statistical period | I/O service time | ms | 10s, 60s, 300s |
DiskUtil | Disk I/O Utilization | Average of I/O utilization of the disk during the statistical period | Percentage of time during which the disk has I/O operations (i.e., non-idle time) | % | 10s, 60s, 300s |
For more information on monitoring metrics, see Cloud Monitor.
Was this page helpful?