Cloud Monitor 2.0 includes the ECS Insights application, which provides comprehensive monitoring for host performance, processes, resources, and events.
Provisioning
Log on to the Cloud Monitor 2.0 console. Click your target workspace. In the Integration Center, two connection types are available for Elastic Compute Service (ECS) observability. Select a connection type based on your requirements:
Elastic Compute Service (ECS): This connection type collects multi-dimensional observable data, such as ECS instance metrics, audit logs, and system events. This lets you connect ECS instances in a region in batches, regardless of their VPC. The collected metrics include agentless metrics from Cloud Monitor and Exporter proxy metrics. The system automatically detects and connects to Linux and Windows instances and installs the Exporter.
GPU Monitoring for ECS: This connection type automatically installs Gpu-exporter on Alibaba Cloud GPU-accelerated instances to collect metrics. Supported operating systems include Alibaba Cloud Linux, CentOS, and Ubuntu.
Select a connection type. On the configuration page, click Determine to use the default configurations, or customize the configuration by following the on-screen instructions. Wait for the connection to complete.
After the connection is successful, you can view it under Integration Management in the Integration Center. A project is then automatically created in Simple Log Service (SLS) to store the metrics and logs.
Application overview
Log on to the Cloud Monitor 2.0 console. Click your target workspace. In the Application Center, click CloudLens for ECS to open the HostOverview page. The overview page summarizes all your Alibaba Cloud ECS instances, including the total number of instances, total vCPUs, total memory, instance health status, and observability administration (the number of connected observable instances).
Account-level instance analysis:
Provides statistics for all Alibaba Cloud ECS instances at the account level, including the total number of instances, total requested vCPUs, and total requested memory.
Provides statistics for all Alibaba Cloud ECS instances at the region level, including the total number of instances, total requested vCPUs, and total requested memory.
Account-level instance status analysis:
Displays the health status of Alibaba Cloud ECS instances. This includes horizontal point-in-time comparisons, horizontal time-series analysis, and statistics on the number of normal and abnormal instances.
Displays the abnormal status of Alibaba Cloud ECS instances (GuestOS). This includes horizontal point-in-time comparisons, horizontal time-series analysis, and statistics on the number of normal and abnormal instances. The GuestOS status is considered abnormal because of an abnormal operation within the user's operating system.
Displays the abnormal status of Alibaba Cloud ECS instances (Host). This includes horizontal point-in-time comparisons, horizontal time-series analysis, and statistics on the number of normal and abnormal instances. The Host status is considered abnormal because of an abnormal operation on the underlying Alibaba Cloud host.
Observability administration (agentless metrics):
Displays the number and percentage of connected instances for agentless metrics (collected from Cloud Monitor), the collection status of Cloud Monitor metrics, and a list of connected instances.
Observability administration (Exporter proxy metrics):
Displays the number and percentage of connected instances for Exporter proxy metrics, the collection status of Exporter metrics, and a list of connected instances.
Observable resource analysis (connected instances):
Resource statistics for connected observable Alibaba Cloud ECS instances, including the number of vCPU cores, requested and remaining memory, requested and remaining disk space, instance type distribution, and process and thread statistics.