All Products
Search
Document Center

ApsaraDB RDS:Custom metrics

Last Updated:Feb 02, 2024

If you want to create a custom monitoring dashboard or compare the statuses of ApsaraDB RDS for SQL Server instances, you can use the monitoring dashboard feature. This feature allows you to specify the RDS instances and metrics that you want to monitor and compare the performance of RDS instances. This provides an overall understanding of the health status of RDS instances.

Background information

Database Autonomy Service (DAS) provides the monitoring dashboard feature for ApsaraDB RDS for SQL Server instances in phases from May 20, 2022.

Create a monitoring dashboard

  1. Log on to the ApsaraDB RDS console.

  2. In the left-side navigation pane, choose Maintenance & Monitoring > Performance Monitoring.

  3. On the Performance Monitoring page, click the Monitoring Dashboard tab.

  4. On the SQL Server tab, click Add Monitoring Dashboard.

  5. In the dialog box that appears, configure the Dashboard Name parameter and click OK.

  6. Click Select Instances and Metrics, select the RDS instances and the metrics that are required, and then click the 选择 icon.

    Note
    • You can select up to 32 RDS instances at a time.

    • For more information about metrics, see Metrics.

    选择实例和指标

  7. Click Confirm.

    Note

    To modify the RDS instances or metrics that you select in the monitoring dashboard, click Add Instances and Metrics.

View the metric trends of an RDS instance in the monitoring dashboard

  1. Log on to the ApsaraDB RDS console.

  2. In the left-side navigation pane, click Performance Center.

  3. On the Performance Center page, click the Monitoring Dashboard tab.

  4. Click the tab for the database engine, select the monitoring dashboard that you want to view, and then specify a time range to view the trend charts of the metrics during the specified time range.

    Note

    When you specify a time range, the end time must be later than the start time, and the interval between the start time and the end time cannot exceed seven days.

    • You can configure the Instance filtering parameter to filter for multiple RDS instances and then view and compare the metrics of the RDS instances.

    • You can turn on Auto Refresh (Every 5 Seconds) for the system to refresh the trend charts of the metrics every 5 seconds.

    • You can turn on Linkage Chart to view the values of different metrics at the same point in time.

    • You can configure Chart Layout to specify the number of metric trend charts that can be displayed in each row.

    • You can click Add Instances and Metrics to modify the RDS instances or metrics in the dashboard.

    • You can click Details in the trend chart of a metric to expand the chart. You can also change the time range to view the changes in the trend of the metric at the specified time range.

    • You can click Delete in the trend chart of a metric to delete the chart from the dashboard.

Metrics

Category

Metric

Description

SQL Server process

cpu_usage

The CPU utilization of the RDS instance within the operating system.

db_io

The number of I/O requests per second.

db_io_read

The number of read I/O requests per second.

db_io_write

The number of write I/O requests per second.

other_size

The amount of disk space occupied by system files

log_size

The amount of disk space occupied by log files.

tmp_size

The amount of disk space occupied by temporary files.

ins_size

The total amount of disk space occupied by the RDS instance.

data_size

The amount of disk space occupied by data files.

Database

qps

The average number of times that SQL statements are executed per second.

connection_reset

The total number of logon attempts from the connection pool per second.

active_temp_tables

The number of active temporary tables.

active_session

The number of active threads.

active_cursors

The number of active cursors.

sessions

The total number of connections.

active_transactions

The number of active transactions.

transactions

The average number of transactions per second.

write_transactions

The average number of write transactions per second.

read_kb

The outbound traffic per second of the RDS instance.

write_kb

The inbound traffic per second of the RDS instance.

Basic monitoring

cache_hit_ratio

The hit ratio of the high-speed cache.

bufferpool

The percentage of pages that are found in the high-speed cache to all pages that are read from disks.

fullscans

The average number of full table scans per second.

autoparam_attempts_per_sec

The number of auto parameterization attempts per second.

forced_parameterizations_per_sec

The number of successful forced parameterizations per second.

sql_compilations

The number of SQL compilations per second.

unsafe_autoparams_per_sec

The number of unsafe auto parameterization attempts per second.

failed_autoparams_per_sec

The number of auto parameterization failures per second.

safe_autoparams_per_sec

The number of safe auto parameterization attempts per second.

resqlcompilations

The number of SQL statement recompilations per second.

lazy_writes

The number of times that dirty pages are written to disks per second.

checkpoint

The number of dirty pages that the checkpoint operation must write per second.

logout

The number of logouts per second.

logins

The number of logons per second.

locktimeout

The number of lock requests that time out per second, including requests for NOWAIT locks.

deadlock

The number of lock requests that resulted in a deadlock per second.

lock_requests_per_sec

The number of new locks and lock conversions per second.

lockwaits

The number of lock requests that the client waits for per second.

lock_waits

The statistics of processes waiting for locks.

latchwaits

The number of latch requests that are not immediately granted per second.

lock_wait_time_ms

The average amount of wait time for each lock request that resulted in a wait.

average_latch_wait_time

The average waiting time to request a latch resource.

table_lock_escalations_per_sec

The number of times that a lock on a table is escalated to the HoBT level or the table level.

average_lock_wait_time

The average waiting time of the requested lock resource.

total_latch_wait_time

The total wait time for locks in the last second.

Note

You can click the 问号 icon to the right of a metric on a dashboard to view the description of the metric.