Elastic Compute Service:View ECS instance monitoring data

Metrics

ECS instance monitoring metrics:

• vCPU utilization: the percentage of allocated compute units in use. A higher percentage indicates a higher vCPU load. You can view vCPU data in the ECS or CloudMonitor console, by calling API operations, or by connecting to the instance directly.

  To view vCPU utilization after connecting to the instance:

  • Windows instance: Open Task Manager and sort processes by vCPU utilization to identify high-consumption processes.

  • Linux instance: Run top and press Shift+P to sort processes by vCPU utilization.

• Network traffic metrics: inbound and outbound bandwidth usage of the instance in Kbit/s.

  ECS monitors public bandwidth. CloudMonitor monitors both public and internal bandwidth. For example, if 1 Mbit/s outbound public bandwidth is allocated and usage reaches 1,024 Kbit/s, the bandwidth is fully utilized.

View monitoring data in the ECS console

1. Go to ECS console - Instances.

2. In the upper-left corner of the page, select a region and resource group.

3. Find the target instance and click its ID. On the Instance Details page, click the Monitoring tab.

4. Specify a time range to view metrics such as vCPU utilization and memory usage.

   Data granularity varies by time range — shorter ranges yield higher resolution. For example, a 1-hour range has a finer aggregation interval than a 6-hour range, producing different average values.

   

   Note

   You can also query monitoring data by calling API operations such as DescribeInstanceMonitorData, DescribeDiskMonitorData, and DescribeEniMonitorData.

The monitoring data available in the ECS console depends on whether the CloudMonitor agent is installed on the instance.

• If the CloudMonitor agent is installed, the ECS console displays operating system metrics (CPU utilization, memory usage, and system load) from the agent, plus basic metrics identical to those in CloudMonitor.

• If the CloudMonitor agent is not installed, the ECS console displays only basic metrics, identical to those in CloudMonitor.

  Note

  Operating system metrics are collected every 15 seconds; basic metrics every 1 minute. See Operating system monitoring.

  The following table lists metrics available without the CloudMonitor agent. Data is collected every 1 minute.

  Metric Name	Description	Unit	MetricName	Dimensions	Statistics
  (ECS) CPU utilization	CPU usage	%	CPUUtilization	userId, instanceId	Maximum, Minimum, Average
  (ECS) Inbound Internet bandwidth (classic network)	Average rate of inbound Internet traffic	bit/s	InternetInRate	userId, instanceId	Maximum, Minimum, Average
  (ECS) Inbound private network bandwidth	Average rate of inbound private network traffic	bit/s	IntranetInRate	userId, instanceId	Maximum, Minimum, Average
  (ECS) Outbound Internet bandwidth (classic network)	Average rate of outbound Internet traffic	bit/s	InternetOutRate	userId, instanceId	Maximum, Minimum, Average
  (ECS) Outbound private network bandwidth	Average rate of outbound private network traffic	bit/s	IntranetOutRate	userId, instanceId	Maximum, Minimum, Average
  (ECS) Read BPS for all disks	Total bytes read from the system disk per second	Byte/s	DiskReadBPS	userId, instanceId	Maximum, Minimum, Average
  (ECS) Write BPS for all disks	Total bytes written to the system disk per second	Byte/s	DiskWriteBPS	userId, instanceId	Maximum, Minimum, Average
  (ECS) Read IOPS for all disks	Read IOPS for all disks	counts/s	DiskReadIOPS	userId, instanceId	Maximum, Minimum, Average
  (ECS) Write IOPS for all disks	Write IOPS for all disks	counts/s	DiskWriteIOPS	userId, instanceId	Average, Minimum, Maximum
  (ECS) Inbound Internet bandwidth by IP address	Inbound Internet bandwidth	bit/s	VPC_PublicIP_InternetInRate	userId, instanceId, ip	Maximum, Minimum, Average
  (ECS) Outbound Internet bandwidth by IP address	Outbound Internet bandwidth	bit/s	VPC_PublicIP_InternetOutRate	userId, instanceId, ip	Maximum, Minimum, Average
  (ECS) Outbound Internet bandwidth utilization by IP address	Outbound Internet bandwidth usage	%	VPC_PublicIP_InternetOutRate_Percent	userId, instanceId, ip	Average
  (ECS) Inbound Internet traffic (classic network)	Inbound Internet traffic	Byte	InternetIn	userId, instanceId	Average, Minimum, Maximum, Sum
  (ECS) Outbound Internet traffic (classic network)	Outbound Internet traffic	Byte	InternetOut	userId, instanceId	Maximum, Minimum, Average

View monitoring data in the CloudMonitor console

1. Log on to the CloudMonitor console.

2. In the left-side navigation pane, click Host Monitoring.

3. (Optional) On the Host Monitoring page, select the target instance. In the lower part of the page, click Batch Install and then click OK.

   If the CloudMonitor agent is not installed, you can install it on the instance. See Install and uninstall the CloudMonitor agent.

   • Without the agent, only basic metrics are available in CloudMonitor.

   • With the agent installed, both basic and operating system metrics are available.

4. On the Host Monitoring page, click the instance ID. View monitoring data on the Basic Monitoring and OS Monitoring tabs.

   Note

   Monitoring data is retained for up to 30 days.