KVCacheStore provides capacity monitoring and performance monitoring to help you understand the capacity usage and read/write access performance of a KVCacheStore. You can configure alert rules for all monitoring metrics in CloudMonitor.
Notes on capacity metrics
Capacity monitoring data reflects real-time values and is not used as the metering basis for billing. KVCacheStore is billed based on the configured capacity. For more detailed metering data, see Query usage.
Notes on performance metrics
Read IOPS and write IOPS are integers. If the number of related requests within one minute is fewer than 60, the monitored value is displayed as 0.
When a chart shows No data, the target KVCacheStore received no client requests during the selected time range.
Monitoring data retention policy
All monitoring charts use a statistical period of 1 minute. Monitoring data is retained for 90 days and then automatically cleared, counted from the time the data is generated. A single query in the console can span up to 30 days.
View monitoring data
Log on to the KVCacheStore console. You can switch the region at the top of the page.
In the left-side navigation pane, choose Monitoring and Auditing > Capacity Monitoring, or Monitoring and Auditing > Performance Monitoring.
From the KVCacheStore drop-down list at the top of the page, select the target KVCacheStore and the time range that you want to view.
The time range can be 1 hour, 3 hours, 6 hours, 1 day, 7 days, 14 days, or 30 days. You can also customize the start date and end date.
View the monitoring charts. The Performance Monitoring page provides two tabs, Performance Monitoring and Client Performance Monitoring, which display the access performance of the entire KVCacheStore and of an individual compute node, respectively.
The capacity monitoring charts in the console are displayed in GiB, while the unit of capacity metrics in CloudMonitor is bytes (B).
Capacity monitoring
Type | Metric | Metric name | Unit | Description |
KVCacheStore | KVCacheStoreCapacity | Total storage capacity | Bytes (B) | The configured capacity of the KVCacheStore. |
KVCacheStoreCapacityUsed | Used capacity | Bytes (B) | The amount of data currently used by the KVCacheStore. | |
KVCacheStoreKeyNumber | Number of keys | Count | The number of keys currently stored in the KVCacheStore. |
Performance monitoring
Type | Metric | Metric name | Unit | Description |
KVCacheStore | KVCacheStoreReadIops | Read IOPS | Count/s | The number of read requests processed by the KVCacheStore per second. |
KVCacheStoreWriteIops | Write IOPS | Count/s | The number of write requests processed by the KVCacheStore per second. | |
KVCacheStoreReadThroughput | Read throughput | Bytes/s (B/s) | The amount of data read from the KVCacheStore per second. | |
KVCacheStoreWriteThroughput | Write throughput | Bytes/s (B/s) | The amount of data written to the KVCacheStore per second. | |
KVCacheStoreReadLatency | Read latency | Microseconds (μs) | The average latency of read requests to the KVCacheStore. | |
KVCacheStoreWriteLatency | Write latency | Microseconds (μs) | The average latency of write requests to the KVCacheStore. | |
Client | KVCacheStoreClientReadIops | Client read IOPS | Count/s | The number of read requests initiated by a single compute node per second. |
KVCacheStoreClientWriteIops | Client write IOPS | Count/s | The number of write requests initiated by a single compute node per second. | |
KVCacheStoreClientReadThroughput | Client read throughput | Bytes/s (B/s) | The amount of data read by a single compute node per second. | |
KVCacheStoreClientWriteThroughput | Client write throughput | Bytes/s (B/s) | The amount of data written by a single compute node per second. | |
KVCacheStoreClientReadLatency | Client read latency | Microseconds (μs) | The average latency of read requests from a single compute node. | |
KVCacheStoreClientWriteLatency | Client write latency | Microseconds (μs) | The average latency of write requests from a single compute node. |
Configure alert rules
You can create alert rules for the capacity and performance metrics of a KVCacheStore in the CloudMonitor console. When a metric meets the alert condition, CloudMonitor sends alert notifications by using the notification methods that correspond to the alert level.
Alert level | Notification method |
Critical | Phone + SMS + Email + DingTalk chatbot |
Warning | SMS + Email + DingTalk chatbot |
Info | Email + DingTalk chatbot |
The alert condition is met when the average value of a monitoring metric over N consecutive statistical periods satisfies the specified threshold condition. You can set the N value separately for each alert level.
Alert conditions vary by metric type. Refer to the actual CloudMonitor console interface for details.