Lessons from Incident Management and Postmortems at Atlassian
summary
Atlassian documents running hundreds of major incidents from 2016–2019 with about 3,000 engineering staff and 250 trained major incident managers. Response uses a four‑point severity scale where sev2 preempts all work, a single decision maker sets severity, and automation (Jira webhooks and OpsGenie features) creates the big red button and comms channels in the first 20 minutes. Teams use chat, video and Confluence for state, templated internal and external updates, role-based rosters, blameless postmortems and strict limits on admin overhead.
AI-written summary · about
chapters (5)
- 0:00Introduction and context
- 3:59Prerequisites for incident success
- 11:18Incident process in action
- 24:35Design principles and conclusion
- 26:18Q&A and resources
related talks
27mManaging Incidents
17mHow Atlassian Unlocks High-Velocity Teams with Jira
53mThe Agile team’s guide to harnessing change and responding to disruption
39mNavigating the Chaos: A Holistic Approach to Incident Management
45mFailure Happens: Improving Incident Response in Large-Scale Organizations
57mThe Curious Incident in your Software in the Day Time
34mIncidents Happen
19mHow to handle incidents at scale