Docs

RDS for SQL Server

Monitoring and alerts

Monitoring metrics

This article introduces the metrics and related instructions for monitoring and alerting support for RDS for SQL Server.

Monitoring indicator description

You can call the GetMetricData interface to query the monitoring metric data of the RDS for SQL Server instance. When calling this interface to query, you need to obtain relevant information about the following monitoring metrics:

Namespace : The product space to which the monitoring indicator to be queried belongs. The product space name for RDS for SQL Server is VCM_RDS_SQLServer.

SubNamespace : The dimension to which the metric to be queried belongs. RDS for SQL Server instances support monitoring metrics for the engine dimension (base dimension).

MetricName : The name of the monitoring indicator to be queried. The MetricName corresponding to each monitoring indicator of the RDS for SQL Server instance is shown in the following table. For more details, please refer to the cloud monitoring indicator query.

Monitoring indicator table

A detailed description of the various monitoring metrics in a SQL Server instance can be found in the following table.

tip

Cloud monitoring acquires data every 30 seconds by default, and displays the collected data after obtaining the average or maximum value according to the display granularity of the query results. For example, when querying QPS data from proxy nodes in the past 3 hours, the cloud monitoring query results are displayed with a granularity of 2 minutes, that is, the data displayed each time is actually the average of the past 4 query results. To view monitoring data, see View monitoring data.

resource monitoring

indicator name	MetricName	unit	indicator description
CPU usage	CpuUtil	%	CPU usage.
memory usage	MemUtil	%	Memory usage.
disk usage	DiskUtil	%	Disk usage space/instance purchase space.
total disk size	DiskTotalSize	GiB	This metric is used to count the total disk size of the measurement object.
disk usage	DiskUsedSize	GiB	This metric is used to measure the disk usage size of the measurement object.
disk average queue length	AvgDiskQueueLength	Count/s	This metric is used to count the number of processes waiting to write to the measured object.
IOPS	IOPS	Count/s	The number of I/O requests processed per second.
network input traffic	NetworkReceiveThroughput	B/s	The number of bytes received per second.
network output traffic	NetworkTransmitThroughput	B/s	The number of bytes sent per second.

Engine monitoring

indicator name	MetricName	unit	indicator description
Number of database connections in use	ConnectionsInUse	Count	The number of connections a user has to the database.
average transactions per second	TPS	Count/s	This metric is used to count the number of transactions initiated per second by the database.
Average batches per second	QPS	Count/s	This metric is used to count the number of Transact-SQL command batches received per second.
logins per second	Logins	Count/s	This metric is used to count the total number of logins launched per second.
cache hit rate	CacheHitRatio	%	This metric is used to count the percentage of pages found in the buffer cache that do not need to be read from disk.
full table scans per second	FullScans	Count/s	This metric is used to count the number of unrestricted full scans per second.
user errors per second	Errors	Count/s	This metric is used to count the number of user errors per second.
latch waits per second	LatchWaits	Count/s	This metric is used to count the number of latch requests per second that were not granted immediately.
lock waits per second	LockWaits	Count/s	This metric is used to count the number of lock requests per second that the caller is required to wait for.
lock requests per second	LockRequests	Count/s	This metric is used to count the number of new locks and lock conversions requested by the lock manager per second.
lock timeouts per second	LockTimeouts	Count/s	This metric is used to count the number of lock requests that time out per second.
deadlocks per second	DeadLocks	Count/s	This metric is used to count the number of lock requests per second that cause deadlocks.
Pages Written Per Second	CheckpointPages	Count/s	This metric is used to count the number of pages flushed to disk per second by a checkpoint or other operation that flushes all dirty pages.
Average SQL compilations per second	SqlCompilations	Count/s	This metric is used to count the number of SQL compilations per second.
Average SQL recompilations per second	SqlRecompilations	Count/s	This metric is used to count the number of statements recompiled per second.
Difference number of primary and secondary node objects	InstanceObjectDifference	Count	The number of object differences between primary and secondary nodes.

Last updated: 2025.10.11 11:45:47