It seems like just yesterday we were using the term “hybrid” cloud and arguing like fishwives over what it meant. The latter half of 2017 saw the rise of the term “multi-cloud” (with many of the same arguments regarding what it meant).
Now, multi-cloud seems far more self-evident to me than hybrid ever did, but in case you were wondering, here you go: it means you’re using multiple clouds to develop, deploy, and deliver applications. Public, private, on-premises, off-premises, cloud interconnects, SaaS… Multi-cloud means you’re taking advantage of all your options to deliver your growing portfolio of applications.
IT cynics – of which there are a preponderance thanks to years of hype cycles and new ‘trends’ – might be leery of this term. Many might view it as little more than the latest marketing madness.
I’m here to tell you, based on the data, it’s not just marketing. It may be madness for those currently managing applications across multiple clouds, but it isn’t just marketing.
To wit, of the more than 3000 respondents to our 2018 State of Application Delivery survey, 87% already operate in a multi-cloud world. That’s on track with the continuing 5% year-over-year increase in organizations that currently plan to deploy more than 25% of their applications in a cloud environment by the end of the year.
That begs the question which cloud environment are they going to choose?
As we noted in our dive into digital transformation trends, organizations are most likely to choose deployment locations on a per-application basis. They’re weighing costs and security and myriad other factors to come up with an answer across all the varying cloud formations. And I do mean all.
Last year we noted that respondents were using an average of 1.8 different cloud models (not providers, just models). This year, that jumped to an average of 2.1 cloud models in use. When we asked specifically what that meant in terms of cloud providers, we discovered that the majority (59%) are already invested in 2-6 different providers. Nearly 1 in 10 (9%) current manage 7-10 different cloud providers.
Respondents are clearly already operating in a multi-cloud world, managing applications across providers and models. In fact, every model showed an increase over 2017. What we’re seeing, though, is a trend toward more public services and more off-premises. On-premises private, for example, gained a mere 1% in use year-over-year, while public IaaS gained 6% and off-premises private grew by a whopping 11%.
It is still true that when respondents are invested in only one cloud model, that model is most likely to be on-premises private, with 51%. But there is growth in other models as the ‘preferred’ cloud and nothing in the data indicates that isn’t a trend likely to continue.
What is still holding back a more explosive embrace of off-premises, public cloud offerings are challenges that need to be addressed. Challenges that exist unsurprisingly around security and performance. A mere 5% of respondents claimed they had no challenges with multi-cloud. Everyone else did. Topping their challenges with 42% of respondents was applying consistent security policies across all applications. A close second was also security related – specifically the challenge of protecting applications from existing and emerging attacks. Forty percent (40%) tagged that as a challenge. 33% are struggling with optimizing the performance of applications. The concern over performance stands out for two reasons: first, because it’s not security. Second, because one of the top impacts of digital transformation was the move to deliver applications from the public cloud. Whether those applications are intended for consumers or corporate use, performance is critical. Slow or unresponsive business-related apps can destroy productivity gains, and it’s even more detrimental for public-facing applications, where the fickle fingers of 8 out of 10 customers are known to have hit ‘delete’ on a poorly performing application.
Interestingly, the number of providers that comprise a multi-cloud environment sometimes appears to have an impact on these challenges.
We see that determining cost-efficiency of various cloud providers becomes less of a challenge as more providers are added. Conversely, visibility into application health (status, performance, and capacity) becomes more of a challenge as an organization uses more providers.
Interestingly, the number of providers in use seems to have little bearing on challenges like optimizing performance and consistent security policies. Optimizing performance was noted as a challenge by 41% of those using 2-6 cloud providers, and 40% of those with 11-20 providers.
These challenges – like their operating environment - are real and need to be addressed, especially as the number of providers in use continues to expand. With different operating models, APIs, consoles, and application services in use, cloud chaos can quickly become the norm and eat away at any operational gains that may have been made by choosing the best cloud for the application.
Multi-cloud is not just marketing. It’s a real thing, with real challenges that real organizations are facing right now.
For more insights on digital transformation, multi-cloud, application services, security, automation and the continuing NetOps transformation, feel free to grab your own copy of our 2018 State of Application Delivery report and follow along on the Twitters with @f5networks and/or the hashtag #soad18. And stay tuned – we’ll have more insights (including data and perspectives not in the report) in forthcoming blogs.