High availability with CloudBees CI on AWS Graviton3 and Amazon EKS
summary
CloudBees CI runs as active/active controllers on Amazon EKS using Amazon EFS for shared storage and Hazelcast synchronization to keep state consistent, claiming rolling restarts, Horizontal Pod Autoscaling and GA availability since September. Deployments use Graviton3 instances (recommended M7g.xlarge) for up to ~20% cost and up to ~60% lower energy, with a default of two replicas that scale out; customers with hundreds or thousands of controllers report reduced downtime and improved throughput for CI/CD workloads.
AI-written summary · about
chapters (7)
- 0:00Introduction and welcome
- 1:50CloudBees CI and Jenkins overview
- 3:36High availability features
- 6:52Problem with traditional architecture
- 11:36Active/active architecture explained
- 14:30Why run on Kubernetes
- 17:21Q&A and closing remarks
related talks
43mRunning CloudBees Core on EKS: Best Practices & Amazon EC2 Spot Instances
20mRunning mission-critical applications on AWS sustainably
30mHow CloudBees CI Brings Enterprise Scale to Jenkins
13mTired of Jenkins downtime disrupting your CI/CD pipelines?
11mAmazon EKS Roadmap
33mOvercoming migration hurdles
51mIncrease Jenkins High Availability in the Enterprise
38mIntroducing Amazon EKS