Pitfalls of Continuous Deployment
summary
Company adopts continuous deployment with Gargoyle feature switches, Vagrant-based dev VMs, Fabric deploys and a single virtualenv per release; Jenkins runs ~30 builds against ~2,000 tests and a full suite that currently takes ~20 minutes. Rollbacks are engineered to complete in under ten seconds; Mule (distributed test runner) proved too unstable. Monitoring uses Graphite + Nagios and Sentry for aggregated exceptions. Concrete pain points: lxml rebuilds, false-positive CI failures, 40k+ LOC, and the plan to enforce coverage on commits.
AI-written summary · about
chapters (13)
- 0:00Introduction to continuous deployment
- 1:24Continuous integration workflow
- 2:27Pros and cons of rapid deployment
- 4:17Environment and infrastructure
- 9:18Managing feature iterations
- 14:35Automated testing with Jenkins
- 18:56Addressing integration issues
- 22:24Optimizing test suites
- 29:12Deploying code to production
- 33:22Managing deployment challenges
- 38:14Database and cache migrations
- 42:23Metrics, logging, and error tracking
- 55:15Closing Q&A
related talks
30mImplementing Continuous Deployment with Jenkins & Gerritt
36mContinuous Deployment with Rails
34mEnabling a Continuous Deployment Mindset
01h 01mA team's transition to Continuous Delivery
34mContinuous Deployment for webapps based on Django
24mContinuous Deployment at Scale
26mHuman Trust in Continuos Deployment
28mClimbing the Mountain of CI/CD: A Tester’s Tale