La virtualisation a mis en évidence les intéractions très fortes entre serveurs, stockage et réseau. Le Private Cloud tente de gommer les frontières entre les silos et tout naturellement on en vient à réfléchir en terme d’organisation avec de nouveaux metiers qui semblent émerger.
Dans l’ article the IT department of Tomorrow paru dans Forbes ces sujets sont évoqués par Elias Khnaser qui est auteur,speaker et practise manager pour la virtualisation chez Artemis Technology.
“As new technologies emerge that reshape the data center landscape, IT management is faced with the task of delegating responsibilities. In today’s IT department, the roles and responsibilities are well-defined and structured. Teams are siloed into very specific verticals. Networking teams handle all aspects of the network. Storage teams handle everything to do with storage; server or systems teams, servers.
With the advent of virtualization, converged infrastructures and cloud computing, this traditional model is no longer valid. All of a sudden, you see systems teams touching networking and becoming a lot more familiar with storage. Networking teams are slowly creeping up on storage functions as Cisco tightens its grip on the fabric switching market
All of this will eventually lead to the dissolution of the traditional siloed teams, and the roles and responsibilities will shift. No longer will we have a storage admin and a network admin and a systems admin. Rather, we will have a datacenter admin. This new role will be the converged role, one that has sole access to the datacenter and is responsible for it from networking to storage to systems. As the concept of Unified Computing gathers more steam and as IT shops everywhere start to look at hardware refreshes, the idea of a converged infrastructure becomes a lot more appealing from a datacenter 2.0 perspective.
Think about how we build our infrastructure today: Network teams rack their own gear, storage teams roll their racks and server teams bring in their stuff. We each rack, stack and cable. We each configure and then communicate back and forth to get all these components to work together. Let’s face it, that is a long, slow and very cumbersome approach. A converged network will roll in servers, storage and networking in a single rack. Connect a few cables, one person or team configures it, and it is ready to join the private cloud and create systems. By converging our IT teams, we can build and expand our datacenter much faster and meet the business needs in a timelier manner.
As we build datacenter 2.0 and our private cloud, computing resources will be maximized and dynamically allocated from different hardware pools. No longer do we buy a server with 2TB of local disk space only to use 500GB. When the Exchange team requires a server, they will no longer go out and buy one, rack it and stack it; instead, they will request one. When the systems team needs a file server with a defined amount of storage, it will be automatically allocated.
The concept of charge-backs also will become crucial. While today charge-backs are a nice thing to have, tomorrow they will become a must-have. Private clouds will be funded from the different departments within an organization based on computing resource consumption.
Today’s model works something like this: The accounting department has a need for a server, so we spec one out, deploy it and charge it back to them. In most cases, however, we just charge them for that server; we do not take into account the space that server needs, the overhead of managing that server and so many other little hidden costs.
Tomorrow’s model will be about figuring out what your computing resources are costing you. It will literally be like a menu with a defined price list. You have your fixed-price menu with different profiles that cater to the different applications. High-intensity applications have a preset profile of resource requirements; low-intensity applications have a different profile. Of course, you will always have the “a la carte” choice, but that comes at a price as well.
Can you avoid this change? Possibly in the short term. Long term, technology will force the change for you, even if you don’t make it. As new types of hardware make their way into the datacenter, the skill sets and focus of these teams start to blur; networking, storage and systems will no longer seem so well-defined. Your staffing and sourcing decisions will change to accommodate the new technologies. Only then will we truly transform IT into a service organization that sustains itself financially and is no longer viewed as a burden on the business but rather as an enabler of business growth.”
Authors
