Deploying SRE Training Best Practices to Production: How We SRE'ed Our SRE Education Program
Talkat SREcon19 Europe/Middle East/Africa · on Thursday, 2019-10-03· at 09:00· Track 2: The Liffey A·
summary
Google’s SRE EDU Orientation replaces lectures with a rigged production system: a breakable photo‑upload app (frontend, logic server, Cloud Storage, Spanner) that instructors trigger via automated silences and pagers. Automation wires onboarding, instructor calendar holds, canaries and timed breakage injection; about 100 trainees per month run exercises. Monitoring shows a big jump in NPS and confidence — a reported 26 percentage points increase in “very likely to recommend” — after moving to hands‑on, SRE‑style testing and post‑mortems.
AI-written summary · about
chapters (16)
- 0:00Intro
- 2:32SREEDU A Brief History
- 4:09Continuum of Training Options
- 6:23Analogy between Software Development and Training Programs
- 7:38How to Apply SRE Principles to a Training Program
- 15:05What Did Our Monitoring Tell Us?
- 15:35Main Goal of SREEDU Onboarding
- 16:45Introduction to SRE EDU Orientation v2
- 18:38v2 Application Requirements
- 21:45Design and Development Challenges
- 25:15Architecture of the Product
- 25:44How Does 2 Work in Practice
- 26:38Automation of Breakages
- 27:52What Does Our Monitoring Tell Us Now?
- 28:54SRE EDU Orientation v2 is Better Instrumented for Observability
- 30:27Instructional Design Principles for Orgs of All Sizes
related talks
11mDeveloper Productivity Engineering
33mLow Context DevOps: Improving SRE Team Culture through Defaults, Documentation
15mBuilding SRE Culture (without SREs, Technically)
52mMaking SDETs from ppl with 0 programming experience
38mQuit Frustrating Your New Developers - Tips From a Teacher
19mDevelopers are all the same
40mCraft: From Concepts to Large-Scale Deployment
48mTeaching Software Engineers to Threat Model: We Did It, and So Can You