Observability in Artifact Registry

This page describes the resources available for administrators to monitor activity and growth in Artifact Registry resources using Google Cloud Observability services.

By default, Artifact Registry is configured to do the following:

  • Send system logs and audit logs to Cloud Logging
  • Send system metrics and resource metrics to Cloud Monitoring
For more information, read the Cloud Monitoring overview and Cloud Logging overview.

Collected logs

Artifact Registry sends the following types of audit log data to Cloud Logging:

Data Access audit logs are disabled by default because these audit logs can be quite large. For more information about Audit logs in Artifact Registry, see Audit Logging.

Query and view logs

You can use following tools to view and analyze your logs:

  • Logs Explorer is designed to help you troubleshoot and analyze the performance of your services and applications. To learn how to view logs in the Logs Explorer, follow the instructions in View logs by using the Logs Explorer.

You can filter for Artifact Registry logs in the Logs Explorer by searching for artifactregistry in the search field.

For more information about querying and analyzing logs, see Query and view logs overview.

For pricing information, see the Cloud Logging sections of the Google Cloud Observability pricing page.

Collected metrics

Artifact Registry sends metric data from the following monitored-resources to Google Cloud Observability.

To view all metrics sent by Artifact Registry to Google Cloud Observability, see Artifact Registry Metrics.

For more information about metrics, read Metrics, time series, and resources.

Artifact Registry project

Metrics that are written against the project resource include request counts and request latency per project for control plane API requests and data plane API requests. Requests to create and modify repositories are control plane API requests, while tools such as Docker, Maven, or pip send requests to the data plane API.

Repository

Metrics that are written against the repository resource include request counts and request latency per repository for control plane API requests and data plane API requests, as well as repository size over time.

What's next