Last month, we hosted our “Doing Cloud Right Webinar” with Jez Humble (DORA CTO, Author) and Anders Wallgren (Electric Cloud CTO). In the webinar, Jez and Anders discuss some of the most striking findings of the recent 2018 Accelerate State of DevOps Report (ASODR), including the fact that organizations that “do cloud right” are 23 times more likely to be elite DevOps performers!
Continue reading for some top takeaways from this insightful webinar.
High Performers Don’t Make Trade Offs
Humble looked back on some of the more interesting findings from the ASODR: “Historically, we found that high performers are not trading off speed for stability, they’re not achieving short lead times and high frequency at the expense of change failure rate time to restore. Instead, we find the high performers do better at all of these things. So, if you only take one thing away from this it should be that high performers are not making trade-offs.”
Wallgren also had an interesting take on high performers and finds that is all starts with how they measure themselves: “The differences between the high and low performers, or even elite performers, is that we don’t measure these things in percentages anymore, we measure them in orders of magnitude. That has a fundamental and profound effect on the ability of organizations to ship software and provide value to their customers.”
Humble also said that being a low-performer has a strong correlation with infrequent releases: “The low performing organizations deploy between once a week and once a month. In terms of their lead time it takes them between one and six months to get those changes from version control into production. They also have these very high change fail rates. Up at the other end, we have our elite performers who are deploying multiple times per day on demand, they can get changes to production in less than an hour.”
Supporting Legacy on Your Journey to Cloud is Key
Wallgren had a very interesting analogy related to the challenges of enterprise businesses finding ways to integrate legacy technology and new:
“We’ve learned a lot in the last 1,000 years about how to build houses. But, we don’t go and tear down the old houses, we just retrofit them. Well, software is the same way — new technologies come along all the time. Docker and containers have taken us a little bit by storm. But that doesn’t mean that mainframes are going away and it also doesn’t mean that the traditional monolithic architectures are going away. We need to figure out in larger IT organizations, how we absorb all these new technologies, how do we take advantage of them, how do we manage and do governance around them.”
Speaking of new and emerging technologies, the National Institute of Standards and Technology defines five essential characteristics of a cloud based operational model: on-demand self-service, broad network access, resource pooling, rapid elasticity, and measured service.
Humble revealed that a minority of organizations are truly utilizing cloud to its fullest extent, but those that do see great success:
“Out of the people who said they were using cloud, only 20% actually agreed they met all five characteristics according to NIST’s definition. However, of that group, those people were 23 times more likely to be elite performers. So it’s hard to actually get cloud right and do it in what’s often called a cloud-native way. But when you do, it really pays off.”
Your Pipeline Needs to Do Cloud Right
According to Wallgren, self-service is one of the most underutilized aspects of leveraging cloud resources, which is a key component to an effective release pipeline platform.
“One of the major situations that we find with our customers and prospects is this idea of ‘I’m waiting for something.’ ‘I’m waiting for an environment,’ or ‘I’m waiting for a build.’ But oftentimes, what they’re waiting for is that build to be deployed or an environment provisioned. That’s a place where providing self-service on-demand access to environments, resources, and tools is very valuable in terms of saving time, cutting down on errors, and being able to repeatedly reliably push things out into new environments.”
Humble also pointed out that it doesn’t necessarily take a specialized platform to win big. “We looked at open source software and we find that using open source software is correlated with high software delivery and organization performance. And indeed, if you’re using open source software, you’re 1.75 times more likely to be in that elite performing group. Those teams are also 1.5 times more likely to expand the open source usage in the future.”
Wallgren, then spoke about the need for accessibility and flexibility to drive success: “We need to be able to see these things and operate them everywhere. And that means not just the product itself. I need to access it on my phone, on my tablet, on my TV, I need to access it all over the place to deliver that to the end customer. We have to orchestrate everything from the very left side of our process all the way through the entire path to production. All of those things in a continuous delivery organization are going to be largely automated and largely non-touch as much as possible.”
Anyone Can Start the Journey Right Now
When it comes to looking for a correlation between success rates and industry, according to Humble there wasn’t much there: “We wanted to see if there was any difference between industries in terms of the ability to achieve high performance and found there wasn’t. You can be in a highly regulated industry like banking, or healthcare, or government, and you’re just as likely to be a high performer as people in other industries. So, while it’s certainly hard to achieve high performance, the fact that you’re working in a regulated industry is not a barrier, you can still do it.”
Wallgren agrees that success (and speed) doesn’t have to be just in lesser-regulated industries: “So often I hear, ‘Well, we’d like to go fast but, we’re regulated, we’re audited, it’s safety, or it’s finance.’ I don’t mean to minimize those concerns, they’re obviously non-trivial concerns for the people in those organizations. But I think we have been kind of trained into this fallacy of you can’t have it both ways.”
For Wallgren and Jez’s complete insights on the Accelerated State of DevOps report, continuous delivery, and doing cloud right, watch the full webinar.
Latest posts by Electric Cloud (see all)
- 10 Do’s of DevOps Webinar with Gene Kim: Key Technical Practices - March 25, 2019
- ElectricFlow 8.5 – Webinar Recap - November 14, 2018
- Doing Cloud Right: Takeaways from Our Recent Jez Humble Webinar - November 6, 2018