Microsoft 365 Outage Root Cause: Metastable Cloud Services Can't Survive the Restart Storm
esrtweet · x · 2026-11-13
Engineer Luke Elin offers an engineering read of the Microsoft 365 outage: traffic wasn't up, capacity was down — servers failing health checks drop out, a secure-connection process gets overwhelmed, and restored servers fall over as soon as traffic returns. Microsoft is reverting a network config change matching the incident start, but changes aren't syncing and some machines are stuck in a bad state. The diagnosis: Microsoft's running cloud services are only metastable and cannot handle the restart storm — a read widely endorsed by fellow engineers.
More from Infra
- bunnie's security chip gets formal RISC-V verification from Harmonic Aristotle — satnam6502 · 2026-11-13
- Nvidia supply chain: 2027 orders unchanged after Huang exit, some run to 2028 — firstadopter · 2026-11-13
- Nvidia leadership shakeup: competitors get a window as buyers rethink single-vendor reliance — DavidLinthicum · 2026-11-13
- Redis creator antirez praises Framework Desktop as his favorite machine for running LLMs — antirez · 2026-11-13
- Attack of the Clouds revisited: what the global Outlook outage says about centralization — technollama · 2026-11-13
- Tencent starts paying domain experts to label AI training data amid compute scarcity — teortaxesTex · 2026-11-13