Your Software is Broken — Pay Attention
summary
Claims concrete numbers: 49% of engineering time goes to finding and fixing bugs and 84% abandon after two crashes. A production-monitoring stack is built around Bugsnag style hooks that collect stack traces and request params, group identical errors, surface user impact and push prioritized alerts into Slack. Workflows include rotas like Bug Warrior rotation, tagging/snoozing, and occasional bankruptcy or focused bug weeks to get back to inbox zero.
AI-written summary · about
chapters (34)
- 0:00Introduction
- 0:27Who am I
- 1:06What is Bug Snag
- 2:03The scary reality
- 2:46What is production monitoring
- 3:34Why do we care
- 3:45Its never been more true
- 4:43Unhappy customers
- 5:40Finding and fixing bugs
- 7:03Production monitoring
- 7:28Pretending nothing is wrong
- 8:38Written tests
- 9:31Waiting for customers to complain
- 10:19Lack of visibility
- 11:08Log statements
- 11:35Lack of ownership
- 12:11Time pressures
- 12:40Sins
- 12:58Principles
- 13:23Acceptance
- 13:59Finding Issues
- 16:14Data Blindness
- 18:11Visibility
- 20:24Prioritization
- 21:55Diagnostics
- 23:55Tooling
- 25:03Collaboration
- 26:20Tracking Progress
- 26:58Team Setup
- 27:11Bug Team
- 28:28Bug Warrior
- 31:39Avoid the sins
- 32:40Take Action
- 33:21Questions
related talks
22mSolving the Drag of Technical Debt with Error Monitoring
34mWhy Good Software Goes Bad
49mObservability in Practice: How to Become Proactive with Bugs
11mWhy we must approach debugging differently
41mDetective Hat: Investigating Production Issues
26mA Practical Taxonomy of Bugs
22mThe Self-Healing Applications Decade | Future of Software Reliability
33mFrom “its broken” to “fix is out”