According to Gartner, between 2009 and 2010, the number of virtual machines in corporate data centres increased more than 50 percent. Yet the traditional data centre network is not optimally designed for server and desktop virtualisation. From the now mainstream and widespread adoption of server virtualisation, companies are already taking the necessary first steps toward deploying a next-generation data centre switching network, one that is more agile and adaptable to the changing needs of the enterprise. Yet many organisations have not been able to reap all of the benefits of server virtualisation because virtual machine (VM) movement requires manual intervention to modify network provisioning – and dealing with this is the next step towards a virtualised network infrastructure.

It is critical that the network, including the data centre, can accommodate and dynamically adapt to the increasingly demanding workload by building the model that Gartner calls “application fluency”. Alcatel-Lucent Enterprise has already adopted an application-fluent approach toward architecting the next-generation data centre network, and it is clear that this is the direction that networks need to be taking.

Virtual Machine movement restricting the network
When a network is made up of multiple VMs, each consisting of an individual operating system and applications, the VMs communicate with each other and to the outside world using a virtual switch. This virtual switch moves networking into the server realm, bringing with it the need to re-deploy traditional network-based tools and solutions for the virtualised environment.

A true data centre fabric will automatically adapt to VM movement to relieve IT of the burden of manually provisioning the network. This way, data centre networks can adapt to the higher bandwidth requirements of media-rich applications such as video while supporting server and desktop virtualisation, and take the next important steps in providing true application fluency across the network.

Monitoring inter-VM communications can be a huge challenge, due to lack of visibility of inter-VM traffic in the network. And as the number of virtual machines on a single server can be scaled from 8-12 VMs today to say 32-64 in the near future, the need to secure virtual machines from external threats, becomes a serious consideration.

The solution: a VEPA
One of the key solutions proposed to address some of these challenges is a VEPA. A VEPA is becoming a real alternative to the virtual switch, and is integral to moving the switching fabric for the data centre into the switch, rather than the server.

A VEPA in effect takes all the traffic generated from virtual machines on a server and moves it out to an external network switch. The external network switch in turn provides connectivity between the virtual machines on the same physical server as well as to the rest of the network infrastructure.

The VEPA therefore makes all VM traffic visible to the external network switch. And by moving VM switching back into the physical network, a VEPA based approach makes existing network tools and processes work consistently across both virtualised and non-virtualised environments.

Application Fluency is key
To enable VEPA within in the switching fabric, the network has to have the concept of Application Fluency to support Virtual Machine motion. Our approach at Alcatel-Lucent focuses on three pre-requisites within the switching fabric to enable applications fluency.

A Low Latency and a lossless Ethernet Data Centre fabric to make sure the VM Motion is transparent between the Switching Layer and the application Server / VM.
Open interworking between the Virtual Machine Hypervisor and the Switching management platform to transparently match the network profile within the switch and the server in terms of switching priority level, security control, VLAN and Access control Lists.
Centralised provisioning and management to automatically provision the virtual network profile to support the VM movement across the entire data centre switching fabric from a single management platform.

VEPA enables a single point of management and control
A VEPA brings network administrative control back to the network administrator, providing a single point of control for provisioning, monitoring, and troubleshooting all virtual machine related networking functions. Offloading the network functions from the server to the network switch also has the benefit of freeing up server resources and making them available for applications, such as video, across the entire enterprise.

As a consequence, the VEPA-based approach has the promise of being able to scale up virtualisation deployments, reduce complexity and cost, and speed up the adoption of virtualisation.

Enabling an open switching market
With server virtualisation gaining broad adoption, complexities of switching traffic between virtual machines both within a server and across servers are increasing. A VEPA-based approach to inter-VM switching provides a compelling alternative to the traditional virtual switch based approach. Standards efforts are already under-way to provide the capabilities needed in the network and server infrastructure to support VEPA, and enable an increasingly open switching market and environment.
 

ShareThis

Tags: Server Virtualization, Storage Virtualization, I/O Virtualization, Network Virtualization, Desktop Virtualization, VMware, Hyper-V, Citrix, RedHat

Read more White Papers»

More VW365 TV»

Read more Web Exclusives»

Keep up to date with the latest industry products, services and technologies from the world's leading IT companies.

• Virtualization Success Depends on Network Automation by Infoblox

• The Infoblox VMware vCenter Orchestrator Plug-in by Infoblox

• Security in the Next-Generation Data Center by Juniper

• Flattening the Data Center Architecture by Juniper

• The Cloud-Ready Data Center Network by Juniper

Latest IT jobs from leading companies.

Click here for full listings»