metrics-server: Features and release notes-Container Compute Service-阿里云帮助中心

metrics-server is a Container Compute Service (ACS) add-on. It collects offline monitoring data from your cluster and exposes the data through the Metrics API. Monitoring data remains viewable through metrics-server even in an offline state. Use the data for Horizontal Pod Autoscaler (HPA) scaling and basic resource monitoring.

Overview

metrics-server is an enhanced version of the open-source Kubernetes metrics-server, adapted for ACS. The add-on lets you view the offline monitoring data of your cluster. The collected data is available through the Metrics API.

For more information about the open-source project, see the official Kubernetes metrics-server documentation.

Instructions

metrics-server is installed by default in ACS clusters. Follow these steps to enable the basic monitoring feature.

Log on to the ACS console. In the left navigation pane, click Clusters.
On the Clusters page, click the name of the target cluster. In the left navigation pane, click Add-ons.
Click the Logs and Monitoring tab, find the metrics-server add-on, and then click Configuration.
In the dialog box that appears, select Enable Kubernetes Monitoring in CloudMonitor and click OK.

Release notes

January 2024

Version	Release date	Changes	Impact
v0.3.9.7-85b3699-aliyun	January 2024	Optimized monitoring performance for large-scale clusters. Optimized the HPA scaling mechanism to ensure smooth scaling even during a rolling update. To use this feature, set the `--enable-hpa-rolling-update-skipped=true` parameter.	This upgrade does not affect your workloads.