This syndicated post originally appeared at Yankee Group Blog » Zeus Kerravala.

One of the most important changes in scaling corporate data center has been the evolution of the application delivery controller (ADC).  The ADC bridges the network and application environments together.  By having the ADC at this strategic control point, organizations are able to scale their data center infrastructure without disrupting the service to the user.

As we move more and applications and infrastructure to the cloud, the ADC will play a similar role in scaling the cloud.  As more traffic moves to and from the cloud, the ADC will play an increasingly important role in ensuring that this can happen without degradation of performance.  That much is well understood.  What’s not well understood is what the form factor of the ADC should be.

Many of the data center appliance vendors have shifted their strategy to build virtual versions.  The thought being that if the server and storage infrastructure is virtual, and then so should be all of the appliances that sit around it.  This would include ADCs, WAN optimizers, security devices and other functions that surround the server infrastructure.

However, before the industry proclaims, “All hail the virtual appliance”, it’s important to understand there are some limitations to the model of the virtual appliance.  With the ADC playing such a key role, it’s critical that cloud providers understand where these limitations are.

I bring this up because over the past month, two of the mainstream ADC suppliers, market leading F5 Networks and niche competitor Citrix, have launched new versions of their ADCs to help scale the cloud.

Citrix has launched a number of virtual editions of its Netscaler ADC and even a platform that uses its Xen hypervisor to run multiple guest ADC instances on a single platform.  This kind of flexibility is ideal for the world of virtualization, where companies are moving virtual workloads around from cloud to cloud.  However, all this flexibility is still bound by the physical limits of the hardware platform.  The Citrix “pay as you grow” model works well when testing cloud services but can be limiting because it based solely on restricting the capability of a simple hardware platform.  If more ADCs are created than the platform can support, there are no options to add to it.

This limitation is the main reason that I like the strategy F5 has taken to scaling the cloud.  Last month F5 released the VIPRION 2400 chassis based ADC designed for the needs of the medium and large enterprise to complement their existing VIPRION 4400 carrier-class chassis. The chassis based solution provides a different level of pay as you grow.  A chassis based ADC allows companies to add more resources as needed through just adding more blades.  This has been common with other parts of the data center, such as network switches or blade servers but is unique to ADCs.  Also, F5 will likely continue to develop faster and more resource rich blades for the VIPRION 2400 chassis, as they have with the 4400, which will allow customers to keep the current platform in place and scale through swapping the blades out.  Again, this is very common with network switches and ADCs are now following.

So while the flexibility of the virtual platforms is attracting lots of industry attention today, we shouldn’t forget the value that good old fashion hardware can provide when scaling the cloud.

The following two tabs change content below.

Zeus Kerravala

Zeus Kerravala is the founder and principal analyst with ZK Research. Kerravala provides a mix of tactical advice to help his clients in the current business climate and long term strategic advice.
Share This Post:
No Interactions

Comments are closed.

Insight and Influence Through Social Media
ZK Research: Home
RSS Feed
ZK Research is proudly powered by WordPress | Entries (RSS) | Comments (RSS) | Custom Theme by The Website Taylor