DevOps Without Measurement is a Fail
summary
Argues that measuring the full application lifecycle is mandatory, not optional, and shows a working stack: New Relic APM, Browser, Synthetics, Infrastructure and Insights with agents, dashboards, service maps, deployment markers and automated rollbacks. Cites scale: 1.4 billion events/minute across 15,000 customers and a surprising finding that 11% of containers <1 minute. Gives results: a transaction improved from 117s to 2s, Trainline links a 300ms delay to $10M loss, and Project Seymour adds ML to surface future problems.
AI-written summary · about
chapters (43)
- 0:00Introduction
- 1:44The Search for Speed
- 3:17Speed Matters
- 4:01Time is Money
- 4:50The Quest for Speed
- 5:57We are changing
- 6:17Everythings dynamic
- 7:54Changing World
- 8:40Dynamic Cloud
- 9:08Customers
- 10:30Dont repeat mistakes
- 10:53Change approach
- 12:01Responsibility
- 13:32Software Measurement Framework
- 14:50Why is it so hard
- 15:01Data sources
- 15:33Simple example
- 17:00Too much data
- 17:22Detect issues
- 17:37Three core requirements
- 20:14What you need
- 21:10What is New Relic
- 23:08Digital Intelligence Platform
- 24:40DevOps Company
- 26:49Project Seymour
- 28:50What does New Relic do
- 29:49How do we help customers
- 31:38Design
- 32:51Development
- 34:27Build
- 35:46Test
- 37:59Test Environment Automation
- 38:45Autoscaling
- 40:02Integration
- 40:41Deployment Metrics
- 41:32Custom Dashboards
- 43:29Health Maps
- 44:06Incident Tracking
- 44:39Recent Events
- 44:54Data Collection
- 46:45Dashboards
- 50:42Black Friday Example
- 58:23Performance Overhead
related talks
33mMeasuring Developer Efficiency in DevOps: Key Questions, Metrics, Best Practices
21mMonitoring Your Apps is not About Devops
25mDevOps: Why do we Keep Messing Up?
25mImplementing Metrics-Driven DevOps: Why and How
29mMeasuring Devops Success with Pipeline Analytics
59mDevSecOps Productivity Metrics and Tools
20mUsing Monitoring to Speed Up Your DevOps Journey
58mDevOps @ Scale — Greek Tragedy in Three Acts