Devops @Scale
summary
A startup grows from 3 to 30 to 100 engineers and the playbook shifts: firefighting with ad hoc rollbacks fails when left-pad 11 lines breaks npm, a bad deploy needs a revert that no longer builds, and a cloud provider restores only a six-hour backup gap. The fixes are concrete: central artifact repo and proxy, Sumo Logic/Splunk logs, Pingdom alerts, Docker/Chef/Puppet pipelines, license/security scanning (Xray), RCA reports and on‑call rotation. Netflix-style all‑hands DevOps rarely scales; dedicated platform team plus non-functional definition of done proves more reliable.
AI-written summary · about
chapters (5)
related talks
48mDevops@Scale
55mDevOps @Scale (Greek Tragedy in 3 Acts)
58mDevOps @ Scale — Greek Tragedy in Three Acts
29mDevOps at Scale is a Hard Problem
18mScaling DevOps in Complex Environments: Challenges, Failures, and Successes
22mOrganising a fast growing DevOps team a real life use case
36mDevOpsSup - Extended DevOps approach
05mHelp! My DevOps team is stuck!