Container Compute Service:kube-scheduler

Introduction to kube-scheduler

kube-scheduler assigns each pod to a suitable virtual node in an ACS cluster based on pod requirements, node resources, and other properties. For background, see the Kubernetes documentation on kube-scheduler.

How it works

kube-scheduler processes pods in a scheduling queue. For each pod, it identifies candidate nodes and selects the best fit using a two-phase pipeline:

1. Filter — each filter plugin eliminates nodes that cannot run the pod (for example, nodes with insufficient CPU or memory).

2. Score — each scoring plugin assigns a score to each remaining node. The final node score is the sum of plugin_score × plugin_weight across all active scoring plugins. The pod is bound to the highest-scoring node.

ACS automatically filters out zones with high resource usage during scheduling. This zone filtering happens in the filter phase and is not affected by scoring plugin weights.

Usage notes

kube-scheduler is installed by default when you create a cluster. For a full list of supported scheduling features, see Scheduling overview.

Configure kube-scheduler

Configure custom parameters for kube-scheduler from the ACS console.

1. Log on to the ACS console. In the left-side navigation pane, click Clusters.

2. On the Clusters page, find the target cluster and click its ID. In the left navigation pane, choose Operations > Add-ons.

3. On the Core Components tab, find the Kube Scheduler component. Click Configuration in the lower-right corner of the card.

4. In the dialog box that appears, configure the parameters described below, then save your changes.

Different scheduler versions support different features and custom parameters. For version-specific feature details, see kube-scheduler.

Change history