Docs

Container Registry

Alerts and notifications

View monitoring information

CR provides runtime monitoring capabilities for instances, helping you track traffic, storage, and request metrics for routine inspections, capacity assessment, and troubleshooting.

warning

This feature is supported only for Standard Edition instances.

Only data from the last 30 days can be queried.

Built-in monitoring metrics reported by CR are free of charge. However, if you actively write custom metric data to the VMP workspace automatically created by CR, you will be billed based on the VMP metric write volume. For more billing information, see Billing overview.

Background

Monitoring metrics

The meanings and purposes of the built-in monitoring metrics in the console are as follows:

Metric	Chart name	VMP metric name	Meaning and purpose
Public Bandwidth (upload)	Public Bandwidth	cr_instance_public_upload_traffic_bytes_total	Traffic generated when clients push image layer data to the CR instance over the public network.
Public Bandwidth (download/transit)			cr_instance_public_download_traffic_bytes_total	Traffic transitted transferred through the CR instance when clients pull container images from it over the public network.
Public Bandwidth (TOS direct download)			cr_instance_tos_public_download_bytes	Traffic generated when clients directly download image layers from TOS after being redirected to its direct download endpoint by CR during image pull operations over the public network.
Prviate bandwidth (upload)	Private Bandwidth	cr_instance_private_upload_traffic_bytes_total	‌Traffic generated when clients push container image layer data to the CR instance over the internal network.
Internal Bandwidth (download/transit)		cr_instance_private_upload_traffic_bytes_total		cr_instance_private_download_traffic_bytes_total	‌Traffic transitted through the CR instance when clients pull container images from it over the internal network.
Private Bandwidth (TOS direct download)			cr_instance_tos_private_download_bytes	Traffic generated when clients directly download image layers from TOS after being redirected to its internal network endpoint by CR during image pull operations over the internal network.
Storage Usage	Storage Usage	cr_instance_storage_usage_bytes	The storage currently consumed by the instance. This information is used to assess storage utilization.
Request QPS	Request QPS	cr_instance_requests_total	‌Total requests per second for the instance. This information is used to monitor whether peak values approach the instance's upper limit.
QPS by API	QPS by API	cr_instance_requests_total		Request QPS by API type. This information helps identify the distribution of request types.

API types

Feature type	Legend item	Description
Image layer requests	blob GET	Requests to download image layer (blob) data. These account for the highest proportion of image pull requests.
		blob HEAD	Requests to check whether image layers exist. These requests are triggered before both pull and push operations.
		blob POST	Requests to initiate image layer uploads.
		blob PUT	Requests to complete uploads of image layer data.
Image manifest requests	manifest GET	Requests for retrieving image manifests to parse versions and layer composition.
		manifest HEAD	Checks whether the image manifest exists.
		manifest PUT	Requests to upload image manifests. These are the last step in an image push operation.
Authentication	token GET	Requests to obtain tokens (authentication requests before pushing or pulling images).
Authentication		token POST	POST requests to exchanges for tokens.
Others	other GET	GET requests other than those listed above. For example, version detection and directory listing.
		other POST	POST requests other than those listed above.
		other PUT	PUT requests other than those listed above.
		other DELETE	Requests to delete image layers, image manifests, tags, or other artifacts.

Prerequisites

You have created a Standard instance. For instructions, see Creating Standard instances or Creating proxy cache registries.

Procedure

Select the target region in the top navigation bar.

In the left navigation pane, click Instance list. Select an entry point depending on your instance type:

Standard instance: Click Standard instance. Find the target instance and click Enter instance.

Proxy repository: Click Proxy cache. Click the target domain name to enter the instance details page.

In the left navigation pane, click Monitoring information.

In the upper-right corner of the page, select a time range and refresh interval for the monitoring information.

Time range: Select the time range for displayed monitoring data. Only data from the ‌last 30 days can be queried.

Refresh interval: Set the refresh interval for monitoring data. You can click the ‌refresh icon‌ to refresh the current monitoring information manually.

What to do next

CR will automatically create a dedicated workspace in VMP named in the format cr-monitor-{Instance name} (for example: cr-monitor-myregistry). When configuring alert rules in the VMP console, select this workspace as the data source. Manual workspace creation is not required.

Configure alert rules: After reporting the instance's monitoring metrics to VMP, you can configure alert rules for these metrics in the VMP console. For instructions on creating alert rules, see Creating alert rules.

View alerts: After you configure an alert rule, if the instance's metrics reach the thresholds specified in the rule, the system will generate an alert. For instructions on viewing alerts, see Viewing alerts.

Last updated: 2026.09.23 14:37:25