How Docker Accelerates Continuous Development at Codefresh
Download as pptx, pdf0 likes195 views
Containers allow developers to instantly run code changes in a full application context for early feature testing and feedback. This allows issues to be discovered and addressed earlier before code is promoted to staging. Only successful code changes that have passed review and testing will be promoted, minimizing risks later in the development cycle. Containers provide complete visibility into the development process through enriched metadata.
How Docker Accelerates Continuous Development at Codefresh
- 1. Containers #101 How Docker Accelerates Continuous Development Oct 2016
- 2. ABOUT ME • Developer, entrepreneur and technology enthusiast • Live in the Silicon Valley since 2008 • Organizer of Containers 101 • Founder & CEO of Codefresh @RazielTabib
- 3. TRADITIONAL CI/CD The first time code changes are reviewed and tested is at the staging phase, when changes are costly to fix
- 4. FEEDBACK IS RECEIVED AND INCORPORATED EARLIER Developers can instantly run their code changes in the full application context for feature preview and usability testing. Feedback then, can be easily incorporated before changes are promoted to staging
- 5. INTEGRATION TESTS FAILURES ARE DISCOVERED EARLIER Run Integration, UI and performance tests on any feature branches using on demand spin up of compositions Failures can then be addressed earlier in the lifecycle before being promoted to staging
- 6. PROMOTE ONLY SUCCESSFUL CODE CHANGES Only changes that have passed review and testing will be promoted, minimizing the risk of costly changes later in the lifecycle
- 7. ENRICH METADATA FOR COMPLETE VISIBILITY
- 8. DEMO
Editor's Notes
- #3: Few words about myself, my name is Raziel, I am a developer and entrepreneur. I live in the bay area for the past 8 years, I am the founder and organizer of Containers 101 meetup and the CEO & founder of Codefresh. My twitter handle is @RazielTabib please follow if you wish to hear announcement about future talks and meetups on the topic of containers and microservices.
- #4: Lets start by talking first about the some of challenges we see in traditional CI. If we are looking at the typical stages of a pipeline, in traditional CI the first time changes are tested for integration or UI as well as the first time they can be demonstrated is at the staging phase. This not only causing a bottleneck of changes waiting to be promoted but also increases the cost of a bug that is found in this stage and impose even further delays.
- #5: What docker allows us to more easily implement is the provisioning, running and sharing our feature branch earlier in the lifecycle. By leveraging docker-machine to provision a host on a remote machine, and spinning up a full application using docker-compose, we can instantly share our feature branch with the team and capture and incorporate feedback before promoting our code changes or making a PR.
- #6: We can also leverage the instant environment to run integration and UI testing earlier in the lifecycle so we can identify breaks and resolve them before making a PR or promoting our code changes.
- #7: Identifying bugs or gaps in functionality early will allow us to promote higher quality code that will less likely to break our integration tests or cause any delays
- #8: Lastly, enriching the docker image metadata, allows us to ensure we promote docker images that have passed all steps in our pipeline successfully.
- #9: Lets see how it works