We had an amazing time at this year’s AWS re:Invent in Vegas. We were able to chat with more than 2,000 developers, operators, and cloud architects that stopped by. A few observations:
- Feature flags were a hot topic. We chatted with quite a few organizations that are already getting value by decoupling deployment from release.
- Observability was a concept echoed across the session agenda and exhibit hall. It gives you the confidence to solve for the unknowns that can haunt deployment processes, production performance, and overall devops quality of life.
Back at the Split booth, we shared how to progressively deliver features. I covered the benefits engineering teams achieve when they manage, monitor, and experiment with features during before, during, and after their release process.
If you weren’t able to make it to re:Invent, or if it was too chaotic when you dropped by, check out the recording below to learn more about how you can deliver features faster.
You can also step through the slides on my SpeakerDeck account.
Commonly Heard at re:Invent
Q: How does Split compare to a feature management platform?
A: Feature management platforms focus on the “Manage” pillar of Progressive Delivery (control with feature flags).
Feature management platforms do a great job of decoupling deployment from release and controlling gradual rollouts, but lack a rigorous statistical engine to address the Monitoring and Experimentation pillars.
Feature flags/feature management is just 1/3 of Split’s Feature Delivery Platform. Our founders saw the full power of platforms at LinkedIn, Microsoft and other high-cadence CD shops and knew that “Monitoring to limit the blast radius” and “Experimenting to prove whether your features have the desired impact” were where the real leverage and power were.
Q: Does Split work for front-end and back-end code?
A: Yes. Split has 10 SDK’s and supports putting the decision making and observation at any layer in a multi-tier solution.
Q: Do you have suggestions for database migrations related to new code rollouts?
A: Yes. Have a look at Chapter 6 in Feature Flag Best Practices (Oreilly)
For Further Listening
If you’re new to many of these concepts, check out our short and snackable six-episode blog/video series, “Safe at Any Speed”, starting with episode one:
If you want a deep dive on Progressive Delivery and the examples from Booking.com, LinkedIn and Facebook of practices that make moving fast, containing the blast radius and learning to validate what works and where further iterations are needed to hit the mark, check out my talk from GOTO Chicago:
Stay up to date
Don’t miss out! Subscribe to our digest to get the latest about feature flags, continuous delivery, experimentation, and more.
With feature flags, you can control the percentage allocation of users you want to be exposed to a specific feature. This process provides risk mitigation and confirms both usability and scalability. Canary releases, or controlled rollouts, serve as an added layer of protection in case something goes wrong. What is…
Every tech company, should be using a robust feature flag system. A well-built system will provide a host of value-adds and efficiencies for your dev team
Feature flagging is a technique development teams deploy to enable easy switches between codepaths in their systems, at runtime. In simpler terms, they’re control structures that toggle on and off the code inside them. Dev teams use feature flags for a wide variety of purposes, from canary releases to A/B…