If the default containerd configuration is insufficient for your workloads, you can customize containerd parameters for a node pool. For example, you can configure multiple registry mirrors for a specific image registry or configure the system to skip certificate validation for an image registry.
Limitations
This feature is available only for node pools that use containerd 1.6.20 or later as the container runtime. If your node pool uses an earlier version, you must upgrade it.
Precautions
Custom containerd settings apply to nodes in batches. These changes take effect immediately on existing nodes and are automatically applied to new nodes.
Do not modify the supported customizable containerd parameters from the command line. Container Service for Kubernetes (ACK) blocks these changes. You must revert any such modifications.
The system automatically corrects non-standard syntax when applying custom containerd settings.
Supported customizable containerd parameters
Custom container runtime configuration
Parameter | Note | Type | Value |
| The maximum number of concurrent image layer downloads. Increasing this value can accelerate pulls for multi-layer images but also increases network and I/O load. | Int |
|
| Specifies whether to ignore anonymous volumes defined in an image with the | Bool | Default value: true |
| The maximum size in bytes of a core file that can be generated when a containerized process core dumps. Key values:
| Int |
|
| The maximum number of file descriptors that can be opened within a container. | Int |
|
| The maximum amount of memory in bytes that a process in a container can lock. | Int |
|
Registry Mirror Configuration
Description | Recommendation |
Configure registry mirrors for an image registry so the container runtime can pull images from alternative registries to accelerate image pulls. Each registry mirror supports custom path mapping through This change does not require restarting containers. |
|
Image Registries Skipping Certificate Authentication (Insecure Registries)
Description | Recommendation |
Allows the container runtime to skip certificate validation for specified image registries when pulling images. This setting is intended for test environments to pull images from registries that use self-signed certificates. This change does not require restarting containers. |
|
Procedure
Modifying containerd configurations does not affect existing containers. To ensure cluster stability, perform this operation during off-peak hours.
Log on to the ACK console. In the left navigation pane, click Clusters.
On the Clusters page, click the name of your cluster. In the left navigation pane, click .
On the Node Pools page, find the desired node pool, and in its Actions column, choose
> Containerd Configuration.Read the precautions, follow the on-screen instructions to add parameters, specify target nodes, and set the batch policy, then click Submit.
See Configuration examples below.
If you remove the custom container runtime configuration, it automatically reverts to the default value.
After you submit, the settings are applied to nodes in batches, which takes time. You can monitor progress and manage the update in the Event Records area, such as pausing, resuming, or canceling it. If a node update fails, troubleshoot the issue and click Continue to retry.
Pausing the update allows you to validate changes on already-updated nodes. When paused, nodes currently being updated will finish, but new updates are postponed until you resume the task. We recommend that you complete the configuration task as soon as possible. Paused tasks are automatically canceled after seven days, deleting all related records and logs.
Configuration examples
Mirror for docker.io | Insecure private registry | HTTP private registry |
In the Registry mirrors section, enter | In the Insecure registries section, enter the private registry address in the Image registry field, in the format | In the Registry mirrors section, enter the private registry address (for example, |
FAQ
Do custom containerd settings affect running services?
No. The changes do not affect running Pods and apply only to Pods created after the update. We recommend that you perform this operation during off-peak hours.
To apply the new settings to existing Pods, you must manually recreate them.
Error with not match XXX or must be between XXX errors
This error usually occurs for the following reasons:
must be between XXX: The value you entered is invalid. Verify that the value is formatted correctly and is within the allowed range.not match XXX: You used an unsupported modification method. ACK blocks direct modifications made to the configuration file from the command line. Revert the parameter to its original value in the file.
Related topics
For more information about the automated O&M capabilities supported by node pools, see Automated O&M capabilities.
If you encounter errors or abnormal behavior with your nodes, Pods, containerd, or kubelet, see Troubleshoot node issues, Troubleshoot Pod issues, and Node and Node Pool FAQ for troubleshooting guidance.