Codefresh Talks Improving Code Velocity Survey Results

Codefresh recently released a survey about improving code velocity.  Dan Garfield, Chief Technology Evangelist of Codefresh, spoke to VMblog about what the survey revealed, and what Codefresh learned from the results.

Dan Garfield: Well, this is the mission of Codefresh. Our goal is to make engineering teams more effective and we do that with our CI/CD platform. We wanted to make sure that we could closely align ourselves with solving that problem by asking our target customers about the challenges they face and the tools they are using.

VMblog: What are some of the biggest revelations that were uncovered by this?

Garfield: There are a couple things that were really confirmed for me. The largest segment of the market was actually not using any CI/CD tools. This confirms what we have thought which is that CI/CD in general is not very accessible or hasn't been because the tooling just hasn't been mature enough and hasn't provided a smooth, simple user experience.

Another thing that stood out is that those who were using CI very rarely have actually automated the entire delivery process. Only 1% of respondents say they have automated the process from code to deployment and this is where we see a huge opportunity for organizations to become a lot more efficient. If you can deliver software just 10, 20, 30 percent faster, the impact on both your own end user's experience, which is getting the latest software faster, and on the productivity of your engineering team, is massive.

VMblog: Can you also translate that into dollars?

Garfield: We actually have customers who do this when they're evaluating Codefresh where they look at what they'll save by adopting an end-to-end CI/CD platform. For a midsize organization we're talking about hundreds of thousands of dollars a year at a minimum after you factor in developer productivity. To give you an example, I talked to someone recently who is implementing a change that was going to make their company and extra $10 million a year. Now, this was a single change and if you average that out over the value of each change and multiply it out across all the teams that are working, making it possible to get things in just a few days faster, we're actually talking about maybe millions of dollars - depending on the size of your organization - in additional business value just from those features being available. Now, there's another tertiary benefit here, which is that if your developers are productive, they're more likely to be happy and to want to stick around.

VMblog: Can you drill down a little bit further about some of the problems that teams face when they're trying to get code to production?

Garfield: The biggest hurdle is a lack of automation. In our survey, we found that 100% of people cited a lack of automation/manual processes as what prevents teams from deploying more often. It really just comes down to having the right tool set and the right practices in order to feel confident in a fully automated code delivery pipeline.

VMblog: How does the Insight provided by survey respondents underscore what Codefresh offers?

Garfield: There's a very direct link between what this survey reveals and what Codefresh offers. Ultimately our goal with our software is to lower the barriers to automation. So, we've designed our platform in such a way that creating pipelines that are robust and reliable, and don't conflict with each other, is very easy. The secret sauce to that is that we have container-based pipelines which means every single step is its own container and has all the benefits of isolation and portability. That actually means that pipeline creation is as easy as linking together several images in a row to do different tasks, and we built a whole library of steps to do just that.

We know traditional CI/CD tools just aren't up to the challenge because the survey respondents identified the lack of automation as their biggest challenge to delivering software faster even when they had a solution in place. Organizations quickly find themselves in operator hell trying to maintain aging infrastructure with a vast library of unreliable plugins. This is a bullseye for Codefesh, it's exactly the problem our Kubernetes-based pipelines solve.

VMblog: Did the survey reveal anything that impacts your market strategy?

Garfield: There are two things that I think we're surprising. One, the popularity of serverless was far ahead of what we expected. In our survey, more than 65 percent said they were actually using serverless already. That's a very high percentage of the market that's actually already starting to use that technology. The other thing, and I think this was really key for us, is that we're seeing a lot of hybrid workloads. Codefresh is obviously very engaged with Kubernetes. We're kind of "Kubernetes First" - a lot of the features that we've developed are things that really help you with using Kubernetes. We found, however, that most people running Kubernetes are also running other kinds of workloads. Usually that's because they have other workload tooling that they're using that is perhaps a better fit than Kubernetes. A good example would be database management. If you've got RDS running, the odds of you replacing that and running Kubernetes are very low. So, what we took away from it was that Codefresh has amazing support for Kubernetes, and we want to make sure we can support your other kinds of workloads too.  We need to emphasize that more and say. ‘Hey, you're using Kubernetes. That's awesome. We have the best tooling for that. But guess what, you can also deploy to serverless. You can also deploy more files if you need to. You can actually build Android applications in Codefresh.' These are things that most people aren't necessarily aware of, but in our survey, it confirmed for us that there are things we need to emphasize and make sure people know.

VMblog: What are the top three takeaways you want readers to understand from the survey results?

Garfield: Missing automation is the number one thing slowing people down, and so it's critical to get that right. The second thing is that there's a massive opportunity to increase your productivity by automating more of the process. The fact that so few organizations were handling automation for everything including the deployment, those organizations tended to deploy much more often and much more successfully. That's something that we should all be striving for is to automate the entire dev and release process. I would say the third takeaway is that the barrier for people using automation has been the accessibility of tools and understanding how easy they are to use. This is something that Codefresh really tries to solve.

Dan Garfield is a full-stack engineer, kubernaut, and has long running experience with Istio and Helm. He's contributed to a number of open source projects. As a Google Developer Expert, and ex-Atlassian, he's travelled the globe teaching and guiding teams adopting Kubernetes. Ask him about his raspberry-pi powered chicken coop.

Source: Vmblog.com

Powered by NewsAPI.org

Keywords:

Technology evangelist • Engineering • Continuous integration • Continuous integration • Tool • Continuous integration • User experience • Systems engineering • Systems engineering • Organization • Software • Software • Productivity • Engineering • Continuous integration • Automation • Automation • Scientific method • Ethics • Tool • Pipeline transport • Insight • Survey methodology • Software • Automation • Pipeline transport • Reliability engineering • Pipeline transport • Pipeline transport • Continuous integration • Tool • Automation • Software • Infrastructure • Plug-in (computing) • Bullseye (comics) • Kubernetes • Kubernetes • Kubernetes • Kubernetes • Kubernetes • Kubernetes • Remote Data Services • Kubernetes • So What (Pink song) • Kubernetes • Hey You (Pink Floyd song) • Kubernetes • Android (operating system) • Critical theory • Rights • Productivity • Automation • Scientific method • Fact • Organization • Automation • Organization • Cognition • Automation • Computer accessibility • Tool • Military engineering • Number • Open-source model • Google Developer Expert • Atlassian • Kubernetes • Raspberry Pi • Chicken coop •