Announcement on CFS Metric Adjustments
To provide you with better monitoring and alerting services, Cloud File System (CFS) has optimized some of its monitor metrics. The optimized monitor metrics will be officially launched at 19:00:00 on April 10, 2024. By then, you will be able to view the latest monitor metrics in the Cloud File System (CFS) Console and Baidu Cloud Monitor (BCM) Console to obtain more accurate monitor metric data.
Adjustment time
April 10, 2024 7:00:00 PM
Adjustment scope
All regions
Adjustment details
| Before optimization | After optimization | ||||
|---|---|---|---|---|---|
| Metric name (Chinese) | Metric name (English) | Unit | Metric name in Chinese | Metric name (English) | Unit |
| Read throughput | ReadThroughput | bytes/min | Read throughput | ReadThroughputNew | bytes/s |
| Read latency | ReadLatency | ms | Read latency | ReadLatencyNew | ms |
| Read io | ReadIO | Times/minute | Read IOPS | ReadIONew | Time/second |
| Write throughput | WriteThroughput | bytes/min | Write throughput | WriteThroughputNew | bytes/s |
| Write latency | WriteLatency | ms | Write latency | WriteLatencyNew | ms |
| Write io | WriteIO | Times/minute | Write IOPS | WriteIONew | Time/second |
| Metadata read io | MetaReadQps | Time/second | Metadata Read QPS | MetaReadQpsNew | Time/second |
| Metadata write io | MetaWriteQps | Time/second | Metadata Write Qps | MetaWriteQpsNew | Time/second |
Impact of adjustment
The optimized monitoring metric values are now more precise. Details:
① The units for read and write throughput have been updated from “bytes/min” to “bytes/s,” and the units for read and write IO have been revised from “times per minute” to “times per second.” Due to these changes in statistical standards, the displayed metric values may decrease.
② The units for metrics such as read latency, write latency, metadata read IO, and metadata write IO remain the same, but their accuracy has been enhanced.
If you have set the above optimized monitor metrics in Baidu Cloud Monitor (BCM), it is recommended that you refer to the Alert Management document and adjust the threshold values of the configured alert rules in a timely manner based on your actual business conditions. This prevents alert failures from causing business losses.