There is an emerging term in the world of cloud computing that has many confused and some excited. It’s called multicloud.
This was the topic of a Mapping Session I participated in last week at GigaOM Structure (you can find a great review of that Mapping Session by Ben Kepes here). The idea of the Mapping Session was to figure out what’s new with the concept of multicloud and identify any new opportunities for the enterprise.
As the name implies, multicloud is a complex composite of cloud resources with potentially multiple public and private elements. This is in contrast to hybrid cloud computing, which is typically taken to mean one public cloud and one private cloud working together.
We are talking about multicloud for a few core reasons:
- It’s inevitable that, by the time that we’ve addressed most of our IT needs by leveraging public and private cloud computing, we’re going to have several of each in the end-state solution.
- Ultimately, we’re headed to a service-based solution where we leverage core services from several public cloud providers, private clouds, and traditional systems that have been service-enabled.
- We now have cloud-governance solutions that allow for the development, deployment, and management of these very complex distributed multicloud solutions.
- The need for tactical and single-purpose cloud-delivered tools, such as design as a service, testing as a service, or management as a service is becoming part of the cloud computing solution for many enterprises, thus adding to the number of clouds the enterprise employs.
- The value from leveraging the “best of breed” technology is becoming more compelling as we understand the proper use of cloud-based resources and the reality that there are no one-size-fits-all clouds.
As Kepes observed,
“In our multi-cloud mapping session, their seemed to be a significant amount of support for a ‘best of breed’ approach where customers cherry pick modular solutions that best meet particular needs. True there are some very real issues to deal with in terms of Data Gravity and the impacts it has on the range of choices that are practical, but this best of breed multi-cloud approach would seem to be a very strong enabler of the more modular, more flexible and more contextual IT architecture of the future.”
The trouble with multicloud is that it’s really not the original cloud computing target as defined to many in traditional IT. Cloud providers sold the enterprise on the single public cloud concept where an AWS or Rackspace subscription would be the ticket to prosperity. However, as enterprises built the first cloud computing test beds, it became clear that they needed more than a single cloud provider, or technology vendor, could offer.
For those with a deep understanding of enterprise IT, this came as no surprise that we would end up, as we have in the past, with an array of technology that has to be integrated together to form a solution that provides the best value to the business. However, the larger value here is perhaps the movement to a service-based architecture, which was another conclusion of the Multicloud Mapping Session.
This means service oriented architecture, or SOA, which has been a tough sell for the past several years as many enterprises viewed it as overly complex and difficult to implement. However, today we command clouds to do our bidding by using APIs or services and mixing and matching these services within composites or processes to create business solutions.
Those are the fundamentals of SOA, or whatever you want to call it, to disconnect from the bad PR that SOA had just a few years ago. Or perhaps it makes a comeback.
The end game here is to onboard sets of services from public and private clouds that provide expandable and cost-effective resources that IT can provision to build and rebuild business solutions. Of course, these services must be managed by using a service governance layer, as well as secured using identity-based mechanisms. Finally, considering that these services are distributed across many different public and private cloud providers, we need more-sophisticated management approaches and technology, which are indeed emerging.
The difficulty around the use of multicloud is that enterprise IT has a less than stellar history when building, deploying, and operating complex distributed and service-oriented systems. There is much enterprise architecture and planning that needs to occur, planning that many enterprises are unable or unwilling to do. Indeed, if these multicloud solutions are poorly architected without carefully considering the core business requirements, history will repeat itself and they will likely fail.
In order to succeed with multicloud we need to initially focus on the architecture and planning, as well as the right technology for the mix. This means enterprises must, perhaps for the first time, become good at creating, deploying, and managing complex distributed systems. If they can pull it off, there is a huge amount of value with multicloud.