What does it take to make multi-cloud work?

It’s over a decade now since the first wave of migration of enterprise IT functions away from the on-prem data centre and into the cloud. In the intervening period, we’ve seen the maturing of the public cloud market and its domination by a handful of hyperscale providers. And we’ve seen cloud adoption by every sort of organisation from major multinationals to SMEs and start-ups.

 

There has also been a well-documented trend towards deploying different workloads to different clouds, according to suitability, as well as the concurrent running of some functions in the public cloud and others within a private cloud. Enterprises can now choose to run their applications and workloads in whichever environment makes the most sense from a cost, performance or functionality point of view.

 

The benefits of multi-cloud are legion: the ability to source best-in-class providers, to choose the most competitive pricing, to enjoy flexibility and scalability. You gain agility, resilience, performance, security not to mention enhanced risk management.

 

But multi-cloud and hybrid cloud strategies come with considerations that must be taken account of. Some of these can only really be described as challenges. Here’s just a few of them:

 

  • Running some functions on-prem and some in the public cloud demands orchestration tools to manage the balance between the two, especially if you also want to dial your public cloud needs up and down according to need. But what’s the best solution of the many available options?
  • These days multi-cloud probably means cloud-native applications built from containers and microservices that use component services from different cloud providers. But how do you tie all this together without heavy added investment?
  • Enterprises are often attracted to a multi-cloud strategy because it saves them from being locked into one cloud provider’s way of doing things. This calls for applications that are portable between different clouds. But portability might mean a certain dumbing-down of functionality, and might come at the expense of being able to fully exploit a cloud provider potential.
  • The hardest type of multi-cloud strategy to get right is the one that isn’t really a strategy at all but a series of disconnected strategies decided at departmental level. So-called ‘shadow IT’ is technology adopted by business units independently of the IT department, which then need to be integrated to give a CIO any idea of what’s going on.

 

The credit column of multi-cloud should outweigh the debit one provided you deploy some solution that takes away management complexity. You will succeed best with orchestration tools that grant visibility into what is going on with all your clouds. And of course you need an on-demand, elastically scalable and highly available multi-cloud network to join the dots. This is course is a challenge all on its own. What should such a network look like, what should its attributes be? And how can it take away complexity across a distributed application landscape?

 

These and other issues will be aired and debated at the following event.

 

And the following links will be good sources for further multi-cloud info:

 

IDC

Cisco

Verizon

Alkira

Aviatrix

VMware

 

 

By Guy Matthews, Editor of NetReporter

Call for Papers

  • About You

  • About Your Paper Submission