Mainframes Aren’t Dead, They’re Just Running Kubernetes Now
summary
Mainframes run Kubernetes by joining S390 workers to x86 clusters (example: Ubuntu on z/VM) and by booting CoreOS/OpenShift on LPARs via HMC with SFTP/HTTP-served ignition files and SAN storage, using Kata/OpenShift Sandbox for confidential containers and IBM Hyperprotect secure enclave for attestation. Key numbers: 90% of credit card transactions still hit mainframes and SLAs target 7 nines availability. Experimentally nodes join the cluster successfully, with a few daemonsets fixed by kernel parameter tweaks.
AI-written summary · about
chapters (11)
- 0:00Modern mainframes intro
- 1:06Why mainframes matter
- 2:55Mainframes vs big servers
- 5:49Hardware advantages
- 7:15Containers on mainframes
- 11:32S390 container challenges
- 13:19Join Kubernetes cluster method 1
- 14:26LPAR and OpenShift method
- 20:33Booting on LPAR
- 22:01Future of mainframe containers
- 23:54Q&A on reliability
related talks
29mWait, People Run Kubernetes on Mainframes?
05mBare Metal to Containers to K8s: How We Got Here
34mPanel: Linux in the Kubernetes Era: Does The OS Still Matter?
23mImplementation Challenges: From HPC to Containers in the Academy
32mPack-em or Stack-em: Advantages of Running OpenShift on Power
21mLessons from the Field: Running Kubernetes at the Wellsite
05mMaking the Impossible Possible with Kubernetes
06mTalking Open Source & Cloud Native