Save Your Pipelines (and Clusters!) with Tekton Results
summary
Tekton Results archives PipelineRun and TaskRun JSON and logs off-cluster so etcd doesn’t fill its 8 GB limit and clusters don’t die, storing records in a Postgres-compatible DB and archives PipelineRun and TaskRun JSON to Postgres with queryable JSON and indexes. A watcher controller sends runs to an API server that streams logs to S3 or PV via gRPC, exposes REST/gRPC queries, supports RBAC and tkn-results, and can auto-delete runs after successful archival so CI scales beyond ~15k objects.
AI-written summary · about
chapters (24)
- 0:00Introduction
- 0:54Show of hands
- 3:17Tekton in Kubernetes
- 5:08Hard Storage Limits
- 6:32Kubernetes CI Example
- 8:07Quotas
- 8:33CI System
- 9:18Tekton Results
- 9:37Tekton Results Components
- 10:49Cel Filtering
- 12:19Query Data
- 12:51Pipeline Runs
- 13:39Task Room
- 14:03Task Run Back
- 14:52Logs
- 15:46Results
- 17:13API Support
- 18:48ReadOnly Role
- 19:37Token Access Reviews
- 20:22Token Permissions
- 20:51Performance
- 25:40Community
- 26:39Questions
- 28:26Logo
related talks
24mWhat a Life! - Understanding the Lifecycle of a Tekton Pipeline
22mLessons Learned Evolving K8s Extensions: The Future Of Tekton Pipelines
37mBatch Processing Tekton Pipelines in OpenShift
47mTekton, a new approach to pipelines
29mCloud Native CD Pipelines with Tekton
14mOperating Tekton Efficiently and Securely for multi-team use cases
49mServerless CI/CD for Kubernetes with Tekton
41mDogfooding Tekton project with Tekton