Link IoT Edge allows you to monitor gateway hosts, drivers, Function Compute resources, scene orchestrations, and other resources.

Prerequisites

  1. An edge instance is created and a gateway is online. For more information, see Set up environments.
  2. Other resources such as drivers, Function Compute resources, and scene rules are allocated to the edge instance. The edge instance is deployed.
  3. The edge instance has permissions to manage CloudMonitor resources. For more information, see Access resources of other Alibaba Cloud services.
  4. Monitoring is enabled for the edge instance. To enable monitoring, turn on the CloudMonitor Status switch on the Monitoring tab of the Instance Details page.

Monitor hosts

After an edge instance is created and its gateway goes online, you can view the metrics of the host on which the gateway runs. To do this, click Host Monitoring on the Monitoring tab of the Instance Details page.

Host Monitoring provides multiple metrics, including the CPU usage, memory usage, disk usage, and data transfer speed. This satisfies your basic requirements on monitoring, operations, administration, and management (OAM) of gateways.

Host Monitoring

Under Host Monitoring, you can select a sampling period for the metrics. The following table describes the available metrics.

Table 1. Metrics
Metric Parameter Description
CPU Usage cpu_total The percentage of consumed CPU resources to the total CPU resources.
Memory Usage memoryusage_utilization The percentage of consumed memory resources to total memory resources.
Average System Load load_1m The average gateway load over the past 1 minute.
load_5m The average gateway load over the past 5 minutes.
load_15m The average gateway load over the past 15 minutes.
Disk Utilization diskusage_utilization The percentage of consumed disk resources to the total disk resources.
Network Inbound and Outbound Rate (Bit/s) networkin_rate The number of bits received by the NIC per second. This is the upstream bandwidth of the NIC.
networkout_rate The number of bits sent by the NIC per second. This is the downstream bandwidth of the NIC.
Network Inbound and Outbound Packets (pps) networkin_errorpackages The number of error packets received by the NIC per second.
networkin_packages The number of data packets received by the NIC per second.
networkout_errorpackages The number of error packets sent by the NIC per second.
networkout_packages The number of data packets sent by the NIC per second.

You can set alerts for Host Monitoring metrics as follows:

  1. On the Monitoring tab, click Alert Configuration.
  2. On the Alarm Rules page of the CloudMonitor console, create an alert rule.
    When you create an alert rule, we recommend that you do not modify the settings in the Related Resource section. For more information about other parameters and how to create an alert rule, see Configure alerts for host monitoring.

In the Instance Processes section, you can view the information of each process that runs on the edge instance. The information includes the CPU usage and memory usage.

Monitor scene orchestrations

Note Only edge instances of the Professional Edition and Standard Edition support monitoring on scene orchestrations.

After assigning a scene rule to an edge instance, you can view the metrics of the scene rule. To do this, click Scene Orchestration on the Monitoring tab, and then click View in the Actions column to the right of the scene rule.

On the page that appears, you can select a sampling period for the metrics Triggers and Exceptions.Scene Orchestration

Monitor edge applications

Note Only edge instances of the Professional Edition and Standard Edition support monitoring on edge applications.

After assigning an edge application to an edge instance and deploying the instance, you can view the metrics of the edge application. To do this, click Edge Applications on the Monitoring tab, and then click View in the Actions column to the right of the edge instance.

On the page that appears, you can select a sampling period for the metrics. The following table describes the available metrics.Edge Applications
Table 2. Metrics
Metric Description
Number of Calls The number of times that the function is called during the sampling period.
Execution Duration The average processing time for requests that are made by the function. Unit: milliseconds.
Number of Exits The number of exits of the function process or container.

Monitor drivers

Note Only edge instances of the Professional Edition and Standard Edition support monitoring on drivers.

After assigning a driver and device to an edge instance and deploying the instance, you can view the metrics of the driver. To do this, click Driver Monitoring on the Monitoring tab, and then click View in the Actions column to the right of the driver.

On the page that appears, you can select a sampling period for the metric Exits.Driver Monitoring