What I Learned
summary
A support org moves from ad hoc fixes to a documented, repeatable process by cataloguing 2,128 incidents over months, splitting teams into incident and problem roles, and introducing triage rotation plus post-incident reviews. Tools are RT early then JIRA for tickets and runbooks (Docker used in platform), not exotic new tooling. Runbooks and short checklists are written first, automated later; outcomes: lower cost per engineer, faster handoffs, and enough improvement that automation becomes feasible after months of adoption.
AI-written summary · about
chapters (12)
- 0:00Introduction and Welcome
- 0:38Audience Interaction and Roles
- 1:28On-Call Experiences and Challenges
- 2:07War Story: Managing Technical Debt
- 2:55Company Background and Evolution
- 4:07DevOps by Default
- 5:30Centralized Support Team
- 6:26Growth and Challenges
- 7:21Implementing ITIL and Process Improvements
- 16:22Runbooks and Incident Management
- 19:58Lessons Learned and Recommendations
- 23:10Q&A Session
related talks
51mThe subtle art of supporting mature products
42mThe subtle art of supporting mature products
16mCreating a Learning Culture
34mCreate a Learning Culture
35mEpistemology of Incidents and Problem Solving
41mAre We Getting Better Yet? Progress Toward Safer Operations
26mMore Culture, More Engineering, Less Duct Tape v3.0
26mMission Critical Mindset Building a Culture of Resilience in DevOps Teams