DevOps - The Good, The Bad, The Ugly
summary
Five years into DevOps at GrubHub, SREs are embedded in stream teams with a dotted reporting line, producing closer engagement and business-level monitoring (orders/sec, fulfillment latency) but creating constant context switching, interrupt-driven incident load, and tug-of-war priorities. Fixes include instrumentation and automation (automated Jira tickets, EC2 login tagging, push-as-code workflows), choosing Python as a common ops language, weekly “Learning Fridays,” and a strong warning to kill the DevOps silo — automation works but audits, production access, and staffing still bite.
AI-written summary · about
chapters (21)
- 0:00Intro
- 3:05The State of DevOps
- 4:03Why DevOps
- 6:52DevOps Org Structures
- 8:19DevOps Org Structure
- 8:52DevOps Benefits
- 12:03DevOps vs Ops
- 12:17Organizational Issues
- 15:01Priorities
- 17:30The DevOps Line
- 21:43Audits
- 25:11The Third Silo
- 27:02Choosing a toolset
- 30:10Addressing skill set gaps
- 32:44Hiring for Ops
- 34:18Training isnt enough
- 36:03Empathy
- 37:22Conclusion
- 38:17Python vs Bash
- 39:15Questions
- 39:47Question
related talks
23mDevOps Is More About Customer Feedback And Quick Learning...
42mDemystifying SRE and DevOps
44mDevOps for developers (or maybe against them?!)- 42mDemystifying SRE and DevOps
31mDevOps for Java Developers (Or Maybe Against Them?!)
22mDevOps XXL: Scale-up in Large Setups with SRE
24mHow DevOps changed the way we operate software
47mDevOps or DevX – Lessons We Learned Shifting Left the Wrong Way