Twitter Facebook Google Plus Linkedin email Print Pdf


Networks in the age of the data center revolution

Posted on 22 August 2011 by Pierre Picard

Jean-François Tourneix

Jean-François took over as Director of the Networks and Security Pre-Sales team in mid-2007, having spent several years with Bull Export developing sales of network solutions and building partnerships with network makers. He is a graduate of the University Paris XIII-Evry, with a degree in Industrial Computing.

With consolidation and virtualization becoming much more widespread, and a major move towards Cloud computing and services-oriented architectures, data centers are currently experiencing an unprecedented upheaval. As an essential component of data center infrastructure, the network is no stranger to this transformation. To achieve the expected benefits ­– in terms of greater operational flexibility, service quality, cost control and a smaller environmental footprint – networks have to adapt to both the technologies and the philosophies of the new-generation data center.

Today’s network: unremarkable, yet critical

In a more and more widely dispersed technical and business environment, the network is used by a constantly growing number of users and carries a highly diverse range of information: from Web applications and telephony, to unified messaging and multi-media data. Meanwhile, both the way we use networks and what we expect from them is changing. As an increasingly proportion of business activities happen on line, the network must operate more and more at the pace of the organization, supporting everything from commercial operations to major projects. For employees and customers alike, the network is in becoming both critical and unremarkable. Users now think it is the most natural thing in the world to have a reliable, fast and permanent connection to all kinds of devices, and to be able to access their data at anytime, from anywhere. Over the past five years, the number of connected systems has increased tenfold, and in 2011 more smartphones with WiFi connections and tablet computers will be sold than PCs; a clear sign of the wholesale shift towards mobility. And the network is in the front line!

However, just as the time when we piled on more and more servers and storage arrays has gone, simply increasing bandwidth cannot be the answer. As a key component of the infrastructure, the network has to adapt to the structural, technological and functional changes that are happening to information systems, but it also has to fit in with the IT Department’s overall approach. In particular, it’s not immune to the need for cost reduction.

Optimization and rationalization

So, the time has come to optimize and rationalize our networks. As much for operational reasons as for technical ones, the network has to follow in the footsteps of server virtualization, and also enable dynamic resource allocation, especially to deal with peaks in work loading. Arranging switches in a cluster is one way to overcome bandwidth, availability and security constraints while avoiding the need to add far too many extra items of equipment. Another emerging and increasingly significant trend is the convergence of SAN and LAN networks on FCoE (Fiber-Channel-over-Ethernet) type infrastructures. Standards for this are gradually being established.

In order to be completely successful, a consolidation approach needs to be closely linked to detailed control of resources. With the emergence of the Cloud and the increasing recourse to on-line services outside the organization, the network will be under even more pressure, because its overall performance is going to depend partly on external suppliers. But from the user’s perspective, the IT Department will still be accountable for service levels. Which means it will have to establish and maintain network governance systems, by putting in place instruments to measure and analyze data flows, and providing precise and regular reports. At an operational level, we are also expecting to see a greater convergence between the various administrative and provisioning tasks, in a context where all infrastructure resources are allocated dynamically.

A landscape in a state of flux

These changes, in terms of approach as well as technology, will see the network becoming increasingly important within the data center, and makers are gearing up for significant growth. Many recent initiatives show how they are hoping to extend their coverage of the data center and IT infrastructure more generally: the launch of Cisco’s UCS servers and, most recently, the Cius tablet; HP’s acquisition of 3Com; the purchase of Foundry Networks by Brocade… Add to this the changing nature of some business functions (for example, Network Managers are increasingly being asked to take on the management of IP telephony, which previously would have been the responsibility of other departments), and the whole network landscape is indeed in a state of total flux.

So from every perspective, the network is at the heart of data center transformation, and changes to the whole function of the IT Department. As a long-standing partner of the main network suppliers, as well as an expert in the technical components of the data center (from servers and virtualization to storage and networks), Bull is well placed to support its customers through this evolutionary process. From up-front consultancy to implementation, right through to network supervision, Bull can bring its multi-faceted experience to bear on the new network landscape. Because the biggest challenges will still be those of optimizing and re-establishing data center performance. And that involves much more than individual components in isolation, so it’s definitely something which has to be tackled globally.

For more information >>>>

Comments are closed.