Feature flag use cases you haven't heard about yet
summary
Feature flags get used here to do real work: wrap logging with a toggle and enable it briefly for a small cohort to find and safely delete dead code, mirror live requests asynchronously to a new service and ramp to 10% to observe load and find breaking points, run parity checkers that compare old and new responses and log mismatches, use mirroring to measure actual spend over a week before switching, and complement with tools like jmeter for synthetic stress tests, one company even left it on for nine months to be sure.
AI-written summary · about
chapters (13)
- 0:00Intro
- 0:49Agenda
- 1:35Feature Flags
- 2:10UI Blocks
- 2:34Use Cases
- 4:02Basic Use Cases
- 4:18Removing Dead Code
- 5:13Scripting
- 5:38Feature Flag
- 6:36Load Stress Testing
- 8:19Evaluating Technical Costs
- 9:10Verifying Parity
- 10:23Conclusion
related talks
56mWhen Feature flags go bad
30mA Tale of Two Feature Flags
08mThe Art of Feature Flagging
19mBeyond the Toggle: Feature Flags as Your Migration Safety Net
01h 09mFeature flags: the toggle, the A/B test and the canary
06mFeature Flag Foul-ups
41mImproving feature flag implementation to achieve flow
41mFeature Flags: Separate Deployment from Feature Release