ARM Migration Made Practical
summary
Explains concrete migration steps: scan repos with Porting Advisor for Graviton (and watch Migration Ease), run realistic dollar‑per‑request performance tests, avoid QEMU emulation and instead add ARM runners to CI and build binaries in parallel, produce multi‑arch images using an image index and populate Dockerfiles with buildkit args, collect config and perf with Aperf, prefer newer runtimes (Java 17/21), then right‑size and optionally use ARM performance libraries for further gains.
AI-written summary · about
chapters (12)
- 0:00ARM migration intro
- 3:11Why ARM matters
- 5:57Where to start migration
- 7:24Managed services to ARM
- 11:38Self-managed software to ARM
- 15:59Custom-built software to ARM
- 21:02Multi-architecture artifacts
- 23:27Container strategies for ARM
- 26:21Reasons not to migrate
- 30:19Technical debt and timing
- 31:22Optimizing after migration
- 33:30Summary
related talks
44mWhat to know before adopting Arm: Lessons learned at Datadog
41mRunning Java on ARM - is it worth it?
33mOvercoming migration hurdles
31mOptimizing Cost and Performance with arm64
01h 14mModernizing and Optimizing Applications for the Cloud
36mWhy and How to Choose Arm64 for your Cloud Applications
52mUbuntu on ARM
31mContinuous Delivery on Arm: A Practical Approach