DevOps Transformation in Large Enterprises
How can large organizations move into the promised land of DevOps that is about fast, consistent and secure automation?
At ChefConf 2019, Damith – Indellient’s VP of DevOps – sat down with Chris from Digital Anarchist to discuss this and the changing landscape of Enterprise DevOps. You can watch the interview over here or read the full transcript at the end of this post.
Throughout their conversation, Chris and Damith openly discussed the DevOps challenges that are unique to large enterprises due to either the scale of their operations and/or their reliance on legacy technologies (ahem, ‘heritage applications’).
Damith highlighted Indellient’s multi-faceted strategy for tackling Enterprise DevOps challenges; we leverage culture, processes and technologies to build a durable solution rather than just focusing on tools and techniques. Damith further talked about the often neglected, yet critical, role of an organization’s culture in determining the success of a DevOps transformation project.
He shed light on why Chef Habitat is a perfect tool for enterprises who want to migrate their legacy application into the cloud and how it enables them to automate their CI/CD pipelines without significantly disrupting their existing toolchain and processes. He mentioned how Indellient’s involvement in one of the largest Chef Habitat deployments to date has made him appreciate the importance of patience and proper planning for DevOps transformation projects.
He further discussed why the horses of the DevOps world (enterprises) should avoid the impulse of implementing Push To Deploy capability (like the Unicorns) from Day One; instead, they should breakdown the project into various stages, identify KPIs for success at every stage and laser focus on hitting those KPIs. Failure to identify and build clear stepping stones in the initial phase of a project may lead to frustration and disillusionment down the road.
Lastly, the conversation touched on the recent trends in the DevOps world; for example, how the looming End Of Life of popular legacy software is pushing companies to initiate DevOps transformation projects and why, all of a sudden, it’s not politically correct to say “legacy software.”
Click here to watch the full video or keep scrolling to read the interview transcript. If you have any questions or concerns about your organizations DevOps project, please contact us by filling out this form.
Chris: Tell me a little bit about your role.
Damith: Sure. I’m from Indellient, and I oversee the DevOps line of business. So we work with SMEs all the way up to the Fortune 20; providing them DevOps solutions both from technology, process and culture perspective
Chris: Oh, culture even, because that’s one of the hardest?
Damith: It is, it is hard it takes a while to get things going and get the culture right. But it’s essential in terms of getting the traction and adoption within an organization; especially, when you think about running operations at scale.
Chirs: So, being on the DevOps side and the DevOps focus, I am assuming you’re dealing with a lot of automation around CI/CD?
Damith: We do. Yeah. We definitely deal with different CI/CD platforms and with different ways to deliver applications from local development to production.
Chris: Yeah. And so, being at ChefConf, you know they launched Habitat version…?
Damith: Habitat 0.81 just came out like a week or so ago. But that’s the latest version of Habitat. And what we see right now is Habitat is really starting to gain some momentum in the marketplace. It’s been around for a couple of years, but the adoption is really starting to pick up now.
Chris: Yeah, so that’s what I wanted to dig into a little bit because I’m familiar with Habitat, but I haven’t yet heard about a full-blown implementation.
Damith: So we’ve been lucky enough to be involved in one of Chefs largest Habitat implementations to date with one of their enterprise customers. And so I think one of the real values – and Berry highlighted this earlier in his discussion – is about legacy applications or heritage applications as they’re called and a lot of the CI/CD solutions today focus on that new or homegrown. But most of the organizations run on a legacy solution and that’s not going to change anytime soon. So it’s really about how do you manage those applications in the same way. You can have the strategy of today, which is just a manual way usually to manage these applications or you can try to find a better way and Habitat provides that better way.
Chris: Somebody I was talking to earlier brought up the heritage as well as that just an easy way to say legacy, I think.
Damith: Yes, I think so too.
Chris: So are you encountering a lot of organizations who are going through that digital transformation, so they’re not starting net new with the CI/CD: they have to look at Heritage applications and bringing them into modern processes?
Damith: Yes, and I think one of the things that have sparked interest with a lot of people is the Windows 2008 end of life. So now they’re like: okay, we have all these applications running on Windows 2008 and end of life is coming up at the end of this year. How are we going to be dealing with this challenge because we can’t rewrite everything, we’ve got to move some stuff up to new platforms and Habitat provides a really good way to help tackle that challenge.
Chris: Yeah, it’s unfortunate in a way that it takes a catalyst like that to really make people go and say, “alright now we need it.”
Damith: So organizations can start there and then they can even look at: okay, now what can I do with my more moderate cloud-native applications and the nice thing with Habitat is it does give you that ONE way to deliver applications to production.
Chirs: So in one of the things that I think a lot of enterprises look at when they start to embrace DevOps is how do they know when they’re doing something correctly. What do they look at to know that they’re actually improving or even taking the first steps into the DevOps world?
Damith: Right. And it all comes down to the ability to quantify correctly and how can organizations quantify improvements. Because if you just go in there and say we need to do DevOps transformation because we really need it but what does that REALLY mean? What are you trying to improve? Where’s your pain? And I think if you don’t start with where the pain is and how you can quantify things; change is really hard to measure. You want to know am I doing this successfully because success for different companies can look very different. Are we going to get too idealistic?: I push a button, and it goes from dev to automated testing to stage and then prod and everything’s live and runs perfectly. Hopefully, one day, everyone will get there. But for most companies, that’s not going to be today. And so it may be cutting down deployments from nine months to like two months. And that’s a significant improvement, and that’s quantifiable. So it’s really important to figure out where the pain point is, how do you quantify it, and then you can figure out what success means for you.
Chris: And organizations need to stop comparing themselves to unicorns.
Damtih: Exactly! You know unicorns, they are cloud-native, they don’t have a lot of legacy.
Damith: Yeah, but that doesn’t mean the enterprises can’t be as Agile as they are: it just might take a bit longer to get that right. And so it’s about taking that iterative approach knowing you’re not going to be 100 percent push-button deployment from day one but having the vision of what does that mean (for you) and how do we get there.
Chris: I mean, does it ever really end.?
Damith: No, I don’t think it does. And because technologies are changing, and you need to make sure your pipelines are resilient to go to whatever the platform of the future is. Today, people are talking about containers and Kubernetes – that’s really hot – but tomorrow who knows where we’ll be? And it could be a different platform. Five, six years ago, we weren’t really talking about Kubernetes the way we are today. It was just emerging, and now it’s the rave right. And so in five years, I don’t know will it be something else, but your applications need to be resilient.
Chris: Absolutely, so what is that one thing that you wish people would stop doing or one of those things that are commonly happening and you’re like, don’t do that anymore.
Damith: I wish people would stop trying to solve all the problems at once. Try to break the problem down into more consumable portions and then bite off something and really look at the problem they’re trying to solve: what does it mean for the business? And to add to that: how are they going to scale it? Because I feel like there are pockets of DevOps teams in enterprises that do really really good work but then when you look at how to scale DevOps outside that pocket and practice it elsewhere, that gets challenging. So just take some time to think through how my solution is going to scale across the enterprise and what does that mean to onboarding new team members, and even internal customers and consumers, onto that platform.
Chris: Well I guess part of that comes from knowing in advance where you want to go, so you’re not just building prototypes.
Damith: Well yeah and that’s hard because when you see that end goal, you try to build for that end goal without building the stepping stones all the time, so it’s hard not to get tempted into like just trying to get there really really quickly. But I think it’s really important to try to build those stepping stones make sure everyone’s involved in the process and scale out and operationalize each one of those phases of your DevOps journey as you go along. So it’s not this massive change that may not be adopted or rejected.
Chris: Damith, thanks for your time. You know it’s always great hearing from somebody who’s been in the field and knows what their customers are experiencing, and I look forward to chatting with you next year and see what you’ve learned at that time in the evolution of your clients. Thanks for your time, as well. Appreciate it.
Want to amp up your DevOps services today? Explore the Pipeline
Marketing & Communications Manager