·E24

Inside Entra Resilience: Microsoft's Outage War Stories, Backup Secrets and Preventing Global Outages

August 23

1h 15m

Episode Description

In this episode, I sit down with my boss, Tarek Dawoud, to pull back the curtain on what really happens during a major service outage.

Tarek shares some incredible "war stories" from his time in the trenches, from the early days of DirSync where the team had to edit a sync file with a debugger to prevent an incident, to the massive outages of 2017 and 2018 that changed everything.

We'll give you a peek into the high-stakes, quick-thinking world of a "live site" incident and reveal the groundbreaking engineering principles like cell-based architecture and the backup authentication service that were born from these challenges, making Entra more resilient than ever before.

Subscribe with your favorite podcast player or watch on YouTube 👇

About Tarek Dawoud

Tarek Dawoud is a Lead Architect in the Customer Engineering team for Microsoft Entra. With years of experience growing up in Entra engineering, he has been involved in his share of outages and has a deep understanding of what it takes to build and maintain a resilient, hyperscale identity service.

LinkedIn - https://www.linkedin.com/in/tarekdawoud/

🔗 Related Links

* SLA performance for Microsoft Entra ID - aka.ms/entraidsla

* Microsoft Blames "Severe Weather" for Azure Cloud Outage

* Microsoft Probes Cause of Global Web Outage

* Microsoft's Azure AD authentication outage: What went wrong

📗 Chapters

00:57 What is a "Live Site"?

14:15 The Secret to Entra's Uptime: Cell-Based Architecture

18:09 How Entra Routes Your Login Request Globally

24:46 War Story #1: The 2017 Conditional Access Outage

29:52 War Story #2: How a Hurricane & an Office Bug Caused Chaos