You can view the historical performance monitoring information of clusters in OceanBase Cloud Platform (OCP).
Performance monitoring for all clusters
Log on to the OCP console.
In the left-side navigation pane, click Clusters to go to the Clusters page.
On the Clusters page, you can view the information about the top 5 clusters sorted by a performance metric. You can specify a performance metric and a time range to view the top 5 clusters sorted by the specified performance metric in descending order.
You can specify a performance metric and a time range to view the information about the top 5 clusters sorted by the specified performance metric. You can click the More icon in the upper-right corner to zoom in the chart for a better view.
The available performance metrics are QPS (times/s), Query Response Time (us), Active Sessions, and CPU Utilization (%).
Metric Description QPS (times/s) The number of SQL statements processed per second in OceanBase clusters managed by the current OCP cluster. Query Response Time (us) The average query response time of the cluster, in microseconds (us). Active Sessions The total number of active sessions in the selected time range. CPU Utilization (%) The average CPU utilization of the cluster. The following time ranges are available: Last 1 Hour, Last 24 Hours, and Last 7 Days.
Performance monitoring information of all tenants
Log on to the OCP console.
In the left-side navigation pane, click Tenants to go to the Tenants page.
In the Top 5 Tenants section on the Tenants page, you can view the top 5 tenants sorted by performance metrics such as TPS, QPS, SQL response time, transaction response time, number of active sessions, event waiting_number of times, event waiting_time, and capacity_number of tables.
You can view the monitoring information for the last 1 hour, the last 24 hours, or the last 7 days. You can click the zoom-in icon in the upper-right corner to zoom in the chart or select a monitoring time range.
The following table describes the performance monitoring metrics.
Metric Description Source QPS The average number of SQL statements processed per second. v$sysstatTPS The average number of transactions processed per second. v$sysstatSQL response time The average time consumed by the server to process an SQL statement, in microseconds (us). v$sysstatTransaction response time The average time consumed by the server to process a transaction, in microseconds (us). v$sysstatActive sessions The number of active sessions. GV$OB_PROCESSLISTEvent waiting_count The number of wait events per second. v$system_eventEvent waiting_time The average time of a wait event, in microseconds (us). v$system_eventCapacity_tables The number of tables. CDB_TABLES
Performance monitoring for a specific cluster
Log on to the OCP console.
In the Clusters list on the Clusters page, find the target cluster and click its name.
In the left-side navigation pane on the page that appears, click Performance Monitoring.
In the Data Filtering area, you can specify the following parameters to filter performance information:
Select Time: This parameter is not displayed in auto refresh mode. Select a time range for the data that you want to query.
Auto Refresh: By default, the auto refresh mode is disabled. After it is enabled, performance monitoring data is refreshed every 5 seconds.
Zone: Select zones to view their performance monitoring data.
OBServer: Select OBServer nodes to view their performance monitoring data.
You can respectively view the performance data of databases and hosts in the cluster on different tabs.
To display other metrics, click Select Metric on the right side and select the desired metrics in the Select Metric panel. The system can display up to 10 metrics.
For more information about database performance metrics and host performance metrics, see the "Monitoring metrics" topic in OCP User Guide of the corresponding version.
Tenant performance monitoring information of a specific cluster
You can view the performance information of the five tenants with the highest workload in a cluster.
Log on to the OCP console.
In the Clusters list on the Clusters page, find the target cluster and click its name.
In the left-side navigation pane of the page that appears, click Tenant Management.
In the Top 5 Tenants area on the Tenant Management page, you can view the top 5 tenants sorted by performance metrics such as TPS, QPS, SQL response time, transaction response time, event waiting_number of times, event waiting_time, capacity_number of tables, and number of sessions.
You can view the monitoring information for the last 1 hour, the last 24 hours, or the last 7 days.
Performance monitoring information of a specific tenant
Log on to the OCP console.
In the left-side navigation pane, click Tenants.
In the Tenants list, click the name of the target tenant to go to the Overview page.
In the left-side navigation pane, click Performance Monitoring.
You can specify the following parameters to filter performance information:
Select Time: This parameter is not displayed in auto refresh mode. Select a time range for the data that you want to query.
Auto Refresh: By default, the auto refresh mode is disabled. After it is enabled, performance monitoring data is refreshed every 5 seconds.
Zone: Select zones to view their performance monitoring data.
OBServer: Select OBServer nodes to view their performance monitoring data.
On the Performance & SQL tab, view the monitoring statistics for the following metrics of the tenant:
- QPS: the average number of SQL statements processed per second.
- Response Time: the response time, in microseconds (us).
- SQL Executions: the number of plans executed per second.
- Waiting for Events: the average number of wait events per second.
- Waiting for Event Time: the average wait time of a wait event, in microseconds (us).
- Request Waiting Queue: the number of SQL statements entering the request queue per second.
- Request Waiting Queue Time-Consuming: the wait time of an SQL request in a queue, in microseconds (us).
- Tenant CPU cost: the CPU utilization of the tenant in a zone or OBServer node.
- MemStore usage percent: the MemStore usage of the tenant in a zone or OBServer node.
- RPC Packet RT: the time used to receive and send RPC messages.
- RPC Packet Throughput: the number of bytes in the RPC messages received and sent in a specified period.
- Sessions: the number of active sessions.
- Memory Usage: the memory usage of the tenant.
- Opened Cursors: the number of opened cursors.
- clog sync delay: the clog synchronization latency in a zone or OBServer node.
On the Transaction tab, view the monitoring data of the following metrics of the tenant:
TPS: the average number of transactions processed per second.
Transaction response time details: the average processing time per transaction on the server, in μs.
Number of Transaction Logs: the number of transaction logs submitted per second.
Transaction Log Volume: the data volume of the transaction logs submitted per second, in bytes.
Transaction Log Time-Consuming: the average time consumed for processing a transaction log on the server, in microseconds (us).
Number of Transaction Lock Waits: the number of transaction lock waits per second.
Average Time to Wait for Transaction Lock: the average wait time per transaction lock (us).
TPS: the number of transactions of the specified type per second.
On the Storage and Cache tab, view the monitoring data of the following metrics of the tenant:
MemStore: the size of memory available for data written into OceanBase Database, in MB.
IOPS: the average number of I/O requests processed per second.
I/O Time: the average time consumed for processing an I/O request, in microseconds (us).
IO Throughput: the average volume of I/O data processed per second, in bytes.
Cache Size: the size of the cache, in megabytes (MB).
Cache Hit Rate: the cache hit rate, in percent (%).
Cache Request Total Count: the number of cache requests.