GitLab metrics and registry features help reduce CI/CD bottlenecks (opens in new tab)
GitLab’s two new beta features target common CI/CD bottlenecks without requiring additional third-party tools. CI/CD Job Performance Metrics provides job-level visibility into duration and failures, while Container Virtual Registry centralizes pulls from multiple registries through a cached GitLab endpoint. Together, they help platform teams identify pipeline problems faster and simplify container management.
CI/CD Job Performance Metrics
- Available in GitLab Premium and Ultimate.
- Limited beta on GitLab.com; available on Self-Managed and Dedicated with ClickHouse configured.
- Adds a job-focused panel to Analyze > CI/CD analytics.
- Shows, for the previous 30 days by default:
- Median (P50) and worst-case (P95) job duration
- Failure rate
- Job name and pipeline stage
- Supports sorting, searching, and pagination to identify slow or unreliable jobs.
- GitLab plans to add stage-level aggregation for build, test, and deploy bottlenecks.
Container Virtual Registry
- Available in GitLab Premium and Ultimate; API-ready in GitLab 18.9.
- Provides one GitLab endpoint for pulling images from multiple upstream registries.
- Supports registries such as Docker Hub, Harbor, Quay, and other sources using long-lived token authentication.
- Uses pull-through caching to:
- Reduce repeated downloads and bandwidth costs
- Improve availability and reliability
- Centralize authentication and registry configuration
- Currently configured through the API, with UI management in development.
- Cloud registries requiring IAM authentication, including Amazon ECR, Google Artifact Registry, and Azure Container Registry, may be supported later.
Beta Access and Feedback
- GitLab.com users can request access through their customer success manager or the feature’s feedback issue.
- Self-managed users can enable the feature flag and configure the virtual registry through the API.
- GitLab is seeking feedback to guide future improvements to both features.
These betas are worth evaluating if your team needs better visibility into pipeline performance or manages images across several registries. The metrics feature can replace custom dashboards, while the virtual registry can reduce registry-related configuration and operational overhead.