Within the first five minutes of his keynote at this year’s AWS re:Invent, Andy Jassy, CEO of Amazon Web Services, said something that made me take notice: “Most ISVs will adapt their software to work on one technology infrastructure platform. Some will do two. Very few will do three.” (These comments start at about the 3:40 mark.)
As I was listening to Mr. Jassy, I recalled reading an interview he did with the Wall Street Journal about two months earlier. In this interview, the WSJ writes that Mr. Jassy said “There will not be only one cloud…but rather a handful of large cloud firms with companies taking a multi-provider approach to the cloud.” Why will companies take a multi-cloud approach? According to this WSJ interview, Mr. Jassy said many customers “worry about getting ‘locked in’ with one infrastructure provider.”
So application developers prefer to write software to a single technology infrastructure platform and might consider two. But companies don’t want to be locked into that single technology infrastructure platform. What to do about that contradiction? I will come back to that at the end of this blog.
It is very unlikely you’ll hear or read anything about “multi-cloud” at AWS re:Invent or any other AWS events. Since the WSJ interview was not an AWS event, Mr. Jassy was probably more free to discuss multi-cloud as a viable option for companies. The CEO of a major cloud provider is unlikely to say “We have the best cloud, but if you want to use the other cloud, that’s OK too” just like the CEO of Ford is unlikely to say “We think the Ford in your garage is the best car, but it’s OK if you want to drive the Toyota parked right next to it.”
So while the major cloud providers aren’t talking much about multi-cloud and you won’t see that term in their marketing literature, those same cloud providers are doing something about multi-cloud.
Think about Microsoft’s acquisition of Cloudyn. Prior to the acquisition, Cloudyn provided tools that worked across AWS, Azure, and Google Cloud Platform “to identify, measure and analyze consumption, enable accountability and forecast future cloud spending.” According to this Microsoft blog and this article by Mitch Wagner in Enterprise Cloud, Cloudyn will “continue to invest in supporting multi-cloud environments including Azure, AWS and GCP” even though it is not a Microsoft product. Why? Because like AWS, Microsoft acknowledges it’s a multi-cloud world.
Not to be left out, Google believes a multi-cloud strategy “can help organizations leverage strengths of different cloud providers and spread critical workloads” based on blog Going multi-cloud with Google Cloud Endpoints and AWS Lambda.
A simple Google search for “multi-cloud” will return pages of results of technology companies big and small, old and new talking about multi-cloud for application portability, cost management, disaster recovery, data management, and more.
So multi-cloud is real and it seems everyone is talking about it – even if they aren’t using the term multi-cloud consistently. (See the video below from AWS re:Invent 2017 in Las Vegas.)
Which brings me back to the contradiction between wanting to write software to a single technology infrastructure platform, but not wanting to be locked into that platform. The approach to this problem that Scality is taking is Zenko – the first multi-cloud data controller. Write to a single API (Amazon S3) and support multiple clouds and file formats (AWS, Azure, Google; object and file) from a single platform. Write once. Support many.
Problem solved for both application developers and the senior business people in their organization who are trying to avoid lock-in.
Interested more in Zenko? Learn more and sign up for a free trial at https://www.zenko.io/.