When you manage multiple database instances, switching between individual instance pages to compare performance metrics is slow and error-prone. Database Autonomy Service (DAS) monitoring dashboards give you a single view to track metrics across instances side by side, correlate anomalies across different metrics at the same time point, and drill into specific charts without leaving the dashboard.
Prerequisites
Before you begin, ensure that you have:
A database instance of one of the following supported types:
ApsaraDB RDS for MySQL, self-managed MySQL, or PolarDB for MySQL
Tair (Redis OSS-compatible) or self-managed Redis
ApsaraDB for MongoDB or self-managed MongoDB
ApsaraDB RDS for PostgreSQL
ApsaraDB RDS for SQL Server
The database instance connected to DAS and in the Accessed state. For connection instructions, see:
Create a monitoring dashboard
Log on to the DAS console.
In the left-side navigation pane, choose Intelligent O&M Center > Monitoring Dashboard.
On the Custom Dashboard page, click a tab to select a database engine.
On the tab for the specified database engine, click Add Monitoring Dashboard.
In the Create Monitoring Dashboard dialog box, set Dashboard Name and click OK.
Click Select Metrics and Instances. In the Select Metrics and Instances dialog box, select the instances and metrics to monitor, then click
.You can select up to 32 instances at a time. For available metrics, see Metrics.
Click Confirm.
To update the instances or metrics in an existing dashboard, click Add Instances and Metrics.
View metric trends
Log on to the DAS console.
In the left-side navigation pane, choose Intelligent O&M Center > Monitoring Dashboard.
On the Custom Dashboard page, click a tab to select a database engine.
Select the monitoring dashboard and specify a time range. Time range constraints:
The end time must be later than the start time.
The interval cannot exceed seven days.
For Tair (Redis OSS-compatible) instances, the interval cannot exceed three hours.
Once the dashboard loads, use the following controls to explore your data:
| Control | Description |
|---|---|
| Filter by Instance | Select specific instances to compare their metrics in the same view |
| Auto Refresh (Every 5Seconds) | Refresh all metric trend charts every five seconds |
| Linkage Chart | Enable multi-metric linkage to view values of different metrics at the same time point |
| Chart Layout | Set the number of metric trend charts displayed per row |
| Add Instances and Metrics | Modify the instances and metrics in the dashboard |
| Details (in a trend chart) | Zoom in on a chart and adjust the time range for that metric |
| Delete (in a trend chart) | Remove a metric from the dashboard |
Tip: Click theicon to the right of a metric to view its description.
Metrics
The following tables list the available metrics for each supported database engine.
ApsaraDB RDS for MySQL and self-managed MySQL
| Category | Metric | Description |
|---|---|---|
| MySQL Server | tps | Transactions per second (TPS) |
| qps | Queries per second (QPS) | |
| active_session | Number of active sessions | |
| total_session | Total number of sessions | |
| delete_ps | Average DELETE statements per second | |
| insert_ps | Average INSERT statements per second | |
| replace_ps | Average REPLACE statements per second | |
| update_ps | Average UPDATE statements per second | |
| select_ps | Average SELECT statements per second | |
| bytes_received | Average bytes received from all clients per second | |
| bytes_sent | Average bytes sent to all clients per second | |
| iops_usage | I/O operations per second (IOPS) usage | |
| InnoDB storage engine | innodb_bp_hit | Read cache hit ratio of the InnoDB buffer pool |
| innodb_bp_dirty_pct | Ratio of dirty pages in the InnoDB buffer pool | |
| innodb_bp_usage_pct | Utilization of the InnoDB buffer pool | |
| innodb_data_written | Average bytes written to the InnoDB table per second | |
| innodb_data_read | Average bytes read from the InnoDB table per second | |
| innodb_rows_deleted | Average rows deleted from the InnoDB table per second | |
| innodb_rows_read | Average rows read from the InnoDB table per second | |
| innodb_rows_inserted | Average rows inserted into the InnoDB table per second | |
| innodb_rows_updated | Average rows updated in the InnoDB table per second | |
| MySQL process | cpu_usage | CPU utilization of the MySQL process. The maximum value is 100% for Alibaba Cloud database instances. |
| mem_usage | Memory usage of the MySQL database within the operating system | |
| iops | IOPS of the MySQL database | |
| disk_usage | Disk usage |
Tair (Redis OSS-compatible) and self-managed Redis
| Category | Metric | Description |
|---|---|---|
| Database | instantaneous_ops_per_sec | Commands processed per second |
| get_qps | Read requests per second | |
| put_qps | Write requests per second | |
| connection_usage | Connection usage | |
| inflow | Data read from the database over the network per second | |
| outflow | Data written to the database over the network per second | |
| used_memory | Total bytes that Redis allocates via its allocator | |
| mem_usage | Memory usage | |
| Redis | avg_rt | Average response time of Redis |
| max_rt | Maximum response time of Redis | |
| proxy.avg_rt | Average response time of the Redis proxy node | |
| proxy.max_rt | Maximum response time of the Redis proxy node | |
| proxy.connection_usage | Connection usage of the Redis proxy | |
| cpu_usage | CPU utilization of the Redis process | |
| Keys | evicted_keys | Keys evicted due to the maximum memory limit |
| total_keys | Total number of keys in the database | |
| expired_keys | Total number of key expiration events | |
| Redis Keys | hit | Number of hit keys |
| miss | Number of missed keys |
ApsaraDB for MongoDB and self-managed MongoDB
| Category | Metric | Description |
|---|---|---|
| MongoDB | opcounters.insert | INSERT statements per second |
| opcounters.query | QUERY statements per second | |
| opcounters.update | UPDATE statements per second | |
| opcounters.delete | DELETE statements per second | |
| opcounters.command | COMMAND statements per second | |
| opcounters.getmore | GETMORE statements per second | |
| connections.current | Number of active sessions | |
| metrics.cursor.open.total | Number of open cursors | |
| metrics.cursor.timedOut | Number of timeout cursors | |
| network.bytesIn | Bytes written to MongoDB per second | |
| network.bytesOut | Bytes read from MongoDB per second | |
| network.numRequests | Requests sent to MongoDB per second | |
| MongoDB process | cpu_usage | CPU utilization of the MongoDB process |
| mem_usage | Memory usage of the MongoDB process | |
| disk_usage | Disk usage of the MongoDB process |
ApsaraDB RDS for PostgreSQL
| Category | Metric | Description |
|---|---|---|
| Host | cpu_sys_ratio | CPU utilization of the PostgreSQL process in kernel state |
| cpu_user_ratio | CPU utilization of the PostgreSQL process in user state | |
| cpu_usage | CPU utilization of the PostgreSQL process | |
| mem_usage | Memory usage of the PostgreSQL process | |
| data.r_s | Read IOPS of the disk on which the data directory resides | |
| data.iops | IOPS of the disk on which the data directory resides | |
| data.w_s | Write IOPS of the disk on which the data directory resides | |
| local_fs_size_total | Total disk space of the host | |
| data.fs.used | Used space of the disk on which the data directory resides | |
| data.fs.usage | Space usage of the disk on which the data directory resides | |
| network_out_io | Outbound network traffic of the PostgreSQL process | |
| network_io | Network traffic of the PostgreSQL process | |
| network_in_io | Inbound network traffic of the PostgreSQL process | |
| PostgreSQL basic monitoring | waiting_connection | Number of pending connections |
| idle_in_transaction | Number of idle sessions | |
| active_session | Number of active connections | |
| total_session | Total number of current connections | |
| tup_inserted | Records inserted per second | |
| tup_deleted | Records deleted per second | |
| tup_updated | Records updated per second | |
| tup_returned | Records returned per second for full table scans | |
| tup_fetched | Records returned per second for index scans | |
| tps | TPS in PostgreSQL | |
| xact_rollback | Transactions rolled back per second | |
| xact_commit | Transactions committed per second | |
| deadlocks | Deadlocks per second | |
| oldest_snapshot | Point in time before which dead tuples can be recycled by the instance | |
| max_sent_delay | Replication latency from the secondary ApsaraDB RDS for PostgreSQL instance | |
| max_replay_delay | Latency at which the secondary instance replays transaction commits | |
| long_query_5s | SQL statements with execution duration >= 5 seconds | |
| long_query_1s | SQL statements with execution duration >= 1 second | |
| long_query_3s | SQL statements with execution duration >= 3 seconds | |
| long_idle_in_transaction_3s | Transactions idle for 3 seconds or longer | |
| long_idle_in_transaction_5s | Transactions idle for 5 seconds or longer | |
| long_idle_in_transaction_1s | Transactions idle for 1 second or longer | |
| long_2pc_1s | Two-phase commit (2PC) transactions lasting 1 second or longer | |
| long_2pc_5s | 2PC transactions lasting 5 seconds or longer | |
| long_2pc_3s | 2PC transactions lasting 3 seconds or longer |
ApsaraDB RDS for SQL Server
| Category | Metric | Description |
|---|---|---|
| SQL Server process | cpu_usage | CPU utilization within the operating system |
| db_io | I/O requests per second | |
| db_io_read | Read I/O requests per second | |
| db_io_write | Write I/O requests per second | |
| other_size | Disk space occupied by system files | |
| log_size | Disk space occupied by log files | |
| tmp_size | Disk space occupied by temporary files | |
| ins_size | Total disk space occupied by the instance | |
| data_size | Disk space occupied by data files | |
| Database | qps | Average SQL statements executed per second |
| connection_reset | Total logon attempts from the connection pool per second | |
| active_temp_tables | Number of active temporary tables | |
| active_session | Number of active threads | |
| active_cursors | Number of active cursors | |
| sessions | Total number of connections | |
| active_transactions | Number of active transactions | |
| transactions | Average transactions per second | |
| write_transactions | Average write transactions per second | |
| read_kb | Average outbound traffic per second | |
| write_kb | Average inbound traffic per second | |
| Basic monitoring | cache_hit_ratio | Hit ratio of the high-speed cache |
| bufferpool | Percentage of pages found in the high-speed cache out of all pages read from disk | |
| fullscans | Average full table scans per second | |
| autoparam_attempts_per_sec | Auto parameterization attempts per second | |
| forced_parameterizations_per_sec | Successful forced parameterizations per second | |
| sql_compilations | SQL compilations per second | |
| unsafe_autoparams_per_sec | Unsafe auto parameterization attempts per second | |
| failed_autoparams_per_sec | Auto parameterization failures per second | |
| safe_autoparams_per_sec | Safe auto parameterization attempts per second | |
| resqlcompilations | SQL statement recompilations per second | |
| lazy_writes | Dirty pages written to disk per second | |
| checkpoint | Dirty pages the checkpoint operation must write per second | |
| logout | Logouts per second | |
| logins | Logons per second | |
| locktimeout | Lock requests that time out per second, including NOWAIT locks | |
| deadlock | Lock requests resulting in a deadlock per second | |
| lock_requests_per_sec | New locks and lock conversions per second | |
| lockwaits | Lock requests the client waits for per second | |
| lock_waits | Statistics of processes waiting for locks | |
| latchwaits | Latch requests not immediately granted per second | |
| lock_wait_time_ms | Average wait time per lock request that results in a wait | |
| average_latch_wait_time | Average wait time to request a latch resource | |
| table_lock_escalations_per_sec | Times a lock on a table is escalated to the HoBT level or table level | |
| average_lock_wait_time | Average wait time of the requested lock resource | |
| total_latch_wait_time | Total wait time for locks in the last second |
PolarDB for MySQL clusters
| Category | Metric | Description |
|---|---|---|
| Database | com_delete | Average DELETE statements per second |
| com_insert | Average INSERT statements per second | |
| com_replace | Average REPLACE statements per second | |
| com_select | Average SELECT statements per second | |
| com_update | Average UPDATE statements per second | |
| qps | QPS of the PolarDB for MySQL cluster | |
| tps | TPS of the PolarDB for MySQL cluster | |
| PolarDB process | cpu_usage | CPU utilization of the PolarDB process. The value 200% indicates that two CPU cores are in use. |
| mem_usage | Memory usage within the operating system | |
| PolarDB Server | active_session | Number of active sessions |
| total_session | Total number of sessions | |
| output_traffic | Average bytes sent to all clients per second | |
| input_traffic | Average bytes received from all clients per second | |
| InnoDB storage engine | innodb_data_read | Average bytes read from the InnoDB table per second |
| innodb_data_written | Average bytes written to the InnoDB table per second | |
| innodb_buffer_dirty_ratio | Ratio of dirty pages in the InnoDB buffer pool | |
| innodb_buffer_read_hit | Read cache hit ratio of the InnoDB buffer pool | |
| innodb_buffer_use_ratio | Utilization of the InnoDB buffer pool | |
| innodb_rows_deleted | Average rows deleted from the InnoDB table per second | |
| innodb_rows_inserted | Average rows inserted into the InnoDB table per second | |
| innodb_rows_read | Average rows read from the InnoDB table per second | |
| innodb_rows_updated | Average rows updated in the InnoDB table per second |