OTA Updates in the Wild: Lessons from the Field
summary
Over-the-air updates need atomic, interruption‑resilient installs with crypto signing, rollback and observability; common choices are a full-image AB scheme (two rootfs copies, so roughly double the storage) or a recovery‑partition tradeoff, and application vs system updates determine reboot needs. Mender offers client+server and Mender supports Zephyr and is industry‑proven; SWUpdate is device‑only and steep learning curve; delta updates can shrink payloads to ~10% of an image but usually require a read‑only rootfs; Hawkbit is a management UI and Hawkbit not actively maintained.
AI-written summary · about
chapters (23)
- 0:00Introduction
- 1:10Why OTA updates matter
- 2:36What's at stake
- 5:45Risks of no OTA
- 6:47Key OTA principles
- 8:33Unreliable network challenges
- 10:00Storage and security limits
- 11:26Safe rollback and observability
- 13:14Versioning and compatibility
- 14:17OTA challenges summary
- 14:39OTA architectures trade-offs
- 15:20AB dual root partitioning
- 17:50Recovery partition approach
- 19:57App vs system updates
- 21:05Full vs delta updates
- 22:09Rollout strategies
- 25:39OTA tooling landscape
- 26:22Mender and SWUpdate overview
- 29:15HawkBit and RAUC tools
- 30:20Balena container platforms
- 31:47Choosing OTA tool
- 32:08OTA best practices
- 35:21Q&A
related talks
56mComparison of Linux Software Update Technologies
39mHow to Choose a Software Update Mechanism for Embedded Linux Devices
42mEvolution of (OTA) Update in the IoT World
40mDevice Management and Delta Update for Embedded Devices with SWUpdate and TUF
31mHow to Leverage an Open Update Protocol to Drive Your Embedded Devices from Any Cloud OTA Provider
34mSafety vs Security: A Tale of Two Updates
52mSurviving in the Wilderness: Integrity Protection and System Update
16mPatchlings: Enabling A/B-Equivalent OTA Updates for Deeply Embedded...