Concourse CI 102
summary
Concourse architecture is explained from pipeline creation to garbage collection with concrete numbers and behaviors: pipelines spawn a pipeline routine that starts five scanner routines for five resource configurations, those scanners create check containers last one hour and reuse resource-type volumes, and build routines create containers from Garden and Baggageclaim using Docker-image volumes. Version targeted is Concourse v5.5, and monitoring examples include Prometheus, InfluxDB and Datadog for metrics and alerts.
AI-written summary · about
chapters (17)
- 0:00Introduction and agenda
- 1:01What is Concourse CI
- 3:08Web node and worker overview
- 4:12Pipeline walkthrough start
- 5:14Unpausing a pipeline
- 6:17Radar and scanner routines
- 9:07Check containers lifecycle
- 11:32Scheduler and build routines
- 13:22Executing job steps
- 15:12Task step execution
- 18:49Job success and garbage collection
- 19:51Handling failed jobs
- 21:15Build tracker and orphan builds
- 23:21Build reaper and log retention
- 26:10Garbage collector process
- 28:38Monitoring Concourse deployments
- 30:03Closing and resources
related talks
46mConcourse: CI/CD, 2020 edition
29mCI That Scales With your Project
44mContinuous Delivery with Concourse CI
29mConcourse in the Real World: A Case Study in CI/CD and DevOps - Greg Meyer & Bryan Kelly, Cerner
37mGetting Started with Concourse CI
01h 32mScaling Your Continuous Deployment using Docker and Containers
53mHow to Continuously Deliver Your Platform with Concourse
33mAn Introduction to BUCC (BOSH, UAA, Credhub and Concourse) - Ruben Koster & Ramon Makkelie
