Single-Function Internal Clouds

The State of Cloud Computing

Single-Function Internal Clouds

By TMCnet Special Guest
Joan Wrabetz, CTO, QualiSystems
  |  July 14, 2015

As a new member of the QualiSystems (News - Alert) team, I have been immersed in many recent discussions with our customers and prospects. In talking to them, it became clear that many are on a journey to IT-as-a-Service. This journey may take a few years, but it will be well worth it.

Given all the industry hype about cloud computing, it’s helpful to look at a model for traveling this journey to IT-as-a-Service. I have spotlighted Stages 2 and 3 for specific emphasis on this path.

Source (News - Alert): RightScale 2014 State of the Cloud Report.

Stage 1: This initial stage is mostly completed for many companies. They have been consolidating their data center resources, standardizing their infrastructure, and virtualizing their applications. If they are using cloud at all, they are accessing public clouds like Amazon Web Services (News - Alert) separately from their IT infrastructures.

Stage 2: In this stage, organizations integrate infrastructure to make management and automation easier. The current slew of converged infrastructure offerings are an example of this trend. IT organizations also virtualize and pool their infrastructure in order to make the allocation of resources and provisioning of applications more elastic. New applications are developed that are designed for cloud enablement. Many don’t even need to be virtualized because they are developed using modern PaaS tools. In order to enable self-service on-demand access for applications in the absence of any internal cloud capability, many companies simply start deploying SaaS (News - Alert) applications. These applications now become cloud silos.

Stage 3:  IT organizations begin to automate both infrastructure and applications, but it appears that the process is still selective. Rather than attempting to automate all of IT, organizations are sensibly tackling the most labor-intensive activities and automating them. This allows organizations to demonstrate the benefits of IT self-service, but it also creates more “cloud silos.”

My prediction for 2015 is that we will see a proliferation of single-function clouds that turn-key IT activities into on-demand services, even with self-service portals in many cases. Each of these cloud silos will incorporate automation. These cloud silos may be implemented using public clouds, by moving select applications to SaaS, or by implementing single-function internal private clouds.

Here are some use cases that we are seeing for these single-function internal clouds:

Lab-Test-as-a-Service – This is an example of an on-demand cloud that takes all of the physical equipment (and virtual infrastructure) in a lab and turns it into a service that can be allocated on-demand to multiple users, letting them run tests inside of that service.

Cyber Ranges – These are the same as Lab-Test-as-a-Service, but where the physical equipment is turned into an on-demand cloud (called a range or arena) that is used to test various security breaches inside of that service.

“Proof of Concept” Clouds – This is an example of turning a specific set of vendor equipment and virtual infrastructure into an on-demand service that can be allocated to customers and prospects for the purposes of running proof-of-concept tests.

Developer/Partner Clouds – This is an example of turning a specific set of vendor equipment and virtual infrastructure into an on-demand service that can be allocated to developers and partners for the purposes of developing applications to the vendor APIs and testing against vendor equipment.

Dev Ops Clouds – This is an example of automating the development and testing process and integrating the steps into a continuous stream. Each step in the automation triggers “on-demand clouds” by either provisioning and starting applications (like configuration management software) or triggering the allocation and configuration of physical and virtual infrastructure to support those applications.

Application/Virtual Machine Provisioning – This is an example of providing a self-service catalog of applications and VMs that are pre-configured so that users can easily provision them on-demand. In many cases, this is being implemented within a single cloud environment such as VMware. 

Stages 4 & 5: All of the prior essential work prepares IT organizations to bear real fruit in Stages 4 and 5. This is where they can expand their use of automation and the amount of their infrastructure that is inside of private clouds. Unfortunately, if IT organizations do not plan for the next stages of maturity, they may end up with many different cloud silos that cannot be easily merged into a single private and hybrid cloud. To avoid ending up in this situation, IT organizations should consider focusing on one or a small number of automation platforms for creating their various cloud silos, both from an application and infrastructure point of view, so that when it comes time to merge them together, the process can be relatively seamless.

It’s important to remember that this journey to the cloud is an ongoing effort, not a single destination. IT managers should plan accordingly to assess what stages they are at now, and what steps will be needed to reach the next desired stage.

Joan Wrabetz is the Chief Technology for QualiSystems. Prior to her current role, she was the Vice President and Chief Technology Officer for the Emerging Product Division of EMC.  She has over 20 years of executive management experience at public and privately held technology companies, has been an executive at a number of startup technology companies, a Venture Partner with BlueStream Ventures (News - Alert), and on the board of directors or advisory board of many early stage technology companies.  Joan holds an MBA from the University of California, Berkeley, an MSEE from Stanford University, and BSEE from Yale University. She holds patents in load balancing, distributed systems and machine learning classification and analytics.




Edited by Maurice Nagle
blog comments powered by Disqus