The Hybrid Cloud Needs to Break Down a Number of Silos to Achieve its Potential
May 02, 2014
The term “silo” is one I’ve come across quite a bit over the past year. The tech industry is pretty fickle when it comes to buzzwords, but this one seems to have some staying power and it is definitely an apt descriptor when it comes to defining the walls and roadblocks that keep popping up as old school, legacy systems and players resist newer trends.
It’s also an appropriate way to describe people – for instance, IT departments are often siloed from management, just as they are siloed (or at least perceive that they are) from systems and processes happening off premises, as is the case with cloud computing. And it is that very phenomenon that appears to be driving the business case for the hybrid cloud, in which some systems and processes remain on premise and in control of IT departments while others are turned over to private and public cloud infrastructure.
In a recent Information Week article, Ziv Kedem, CEO of Zerto, a provider of business continuity and disaster recovery solutions for the cloud, wrote about workload and data center silos and how hybrid cloud solutions have the potential to break them down, offering massive advantages.
Typically, workloads in a data center are siloed by a hypervisor and cannot move among different hypervisors easily. Similarly, and particularly in legacy on-premises infrastructure, workloads are siloed by hardware including vendor lock-in to a specific brand. This also creates difficulties when it comes to moving and migrating workloads and applications.
Workloads can also be siloed in clouds, particularly without shared management of resources across disparate cloud platforms. When cloud providers use proprietary infrastructure and tools, it becomes a huge hassle to try to integrate with other clouds and providers, leading customers to become “locked in” to one cloud provider.
The hybrid cloud attempts to break down these silos, offering a blend of in-house and cloud infrastructures that, theoretically, work harmoniously. Not only does the hybrid cloud break down workload silos, but it addresses the concerns of IT departments, which are reluctant to hand over carte blanche control of all their infrastructure and processes to a cloud provider.
According to Kedem, a cloud fabric layer is the key component to bridging the gap to the cloud, eliminating those silos and making hybrid cloud computing successful. More specifically, by paying careful attention to the development of the infrastructure layer, enterprises and cloud providers alike can benefit from interoperability of data and applications across disparate hypervisors, networks and hardware.
A cloud fabric layer should consist of a cross-hypervisor and hardware agnostic transport layer for all data and applications and should also enable orchestration of mobility for complex apps. It will also handle all application dependencies like IP configuration, and will feature production-level tools that enable high service levels of data mobility as well as protection. With those components in place, disparate on premises, private and public cloud infrastructure will interoperate smoothly, creating an optimum hybrid cloud setup.
With Gartner (News
- Alert) estimating nearly half of large enterprises will have some form of hybrid cloud infrastructure in place by 2017, it’s clear that this trend is gaining momentum. But to truly take advantage of the cost savings and efficiencies of hybridization, enterprises and cloud providers are going to have to break down their silos, play nice, and figure out the best way to integrate their infrastructures.
Edited by Rory J. Thompson
Article comments powered by