CBR spoke to Derek Collison, CEO, Apcera about the challenge of running multiple clouds and trust.
There has been plenty of discussion about the use of multiple clouds, Collison said the main reasons for wanting this is: "The front tier for why you want multi cloud is the class of services, and I think there are three main puppets.
"Very high end data services, Amazon has a good needle there, but obviously Big Data is putting constraints on that. Human machine interfacing, like Alexa out of Amazon and the last big one is Big Data/machine learning AI."
The appeal of Big Data and a machine learning AI, is what Collison believes was the reason why Amazon lost a client to IBM.
"I could be wrong here, there might be more, might be less, but my understanding is that the only big client that ever left Amazon, went to IBM because they wanted the app to access Watson.
"It wasn’t because of cost, it wasn’t because of geo footprint, it wasn’t because of disaster recovery but they actually made the switch.
"My understanding is they had to bring everybody, so they might have only had on app that really needed it to drive business value, but they were like crap, we don’t want to hire more people, we’re going to take the pain."
Content from our partners
Not hiring more people is an important factor, he said: "The only thing that is getting more expensive in IT is people."
However, the complexity of different factors with cloud such as how to provision, how to secure and how to network together is forcing businesses to hire more people.
Collison, said: "We need to hire people not only to port the workload to this cloud provider but we actually need a team to actually understand how to stand up that cloud."
Building trust is an important element of technology, with security challenges all around and new technology coming out on a daily basis, building trust is a challenge.
Part of the complexity is that it is impossible to implement policies at scale for a developer.
He said: "You have to understand what version of TLS, data sovereignty rules, where you’re allowed to run what you’re allowed to do.
"We know that doesn’t scale because we’ve seen more entrenched fortune 500 companies try to start embracing that model and not understand that, they got their name in the papers, in a bad way."
To solve this, Collison says the right direction is: "Anything that starts moving in that direction where the platform takes on kind of the execution engine of melding very fast changing and always migrating policies with new versions of apps, new services, scale up scale down.
"I don’t think I’m off 180 degrees, but it’s going to be somewhere in this and no one is getting it, they’re solving the how to we deploy faster, how do we do a PaaS on Amazon or Google now or OpenStack how do we replace VMware.
"At the end of the day all those pieces play a role but it’s to get to some sort of pool of resources that you can trust and say please do this for me."
