Rackspace Teams With Akamai; Could Be Good News for Limelight
Last week, Rackspace announced that it was partnering with Akamai for web acceleration and cloud optimization, and some well-placed pundit commentary suggested that it would be at the expense of Rackspace's relationship with Limelight.
On Friday morning I noticed that Limelight's shares had dipped a little while all the other tech stocks I track had risen. I couldn't immediately see why and had assumed it was just one of those unexplainable fluctuations that so often happens.
At lunch, however, a colleague asked me if i had seen the Rackspace/Akamai news. He and I are working on a large multi-cloud/multi-CDN virtualization project, and this news-which I hadn't yet seen-had some direct bearing on our planning.
After lunch I hit the search engines and also scanned my email. As a customer of both Akamai and Rackspace, I expected the news feeds to reveal something about this news, and indeed Rackspace had made an announcement confirming what my colleague had mentioned.
Now its no news to readers of this site that cloud infrastructures form partnerships with CDNs. However Rackspace has, in my mind done, something very strange here: There are a number of reports quoting "sources at Rackspace" that claim that the relationship between Rackspace and Limelight (that has existed for several years) will be "phased out over time."
In today's marketplace, and from the perspective of a relatively experienced implementor of both CDN and cloud infrastructures (particularly with regard to streaming media deployments), I think this posturing of leaving one partner to run into the hills with another is actually negative for Rackspace. Now I have to add that I can actually find little to show that the claims to "phase out" the Limelight relationship are actually a verifiable strategy of Rackspace's, so I want to highlight that before I make my next statements. If it is indeed the case then It implies that they are moving away from a multi-vendor resilience strategy to wholly embrace a single relationship with Akamai. In my mind that is like throwing the parachutes out of the airplane: You would have to have a really good reason to do so, since it would in nearly every conceivable situation make absolutely no sense at all.
It implies to me that there is some larger deal-making at stake going on behind the scenes.
I also notice that a lot of the reports mention the volumetrics of servers managed by both Rackspace and by Akamai-"one runs X may tens of thousands of servers, and the other runs Y many," etc.
Frankly, I am tired of hearing Akamai sustain its market proposition simply by sheer brawn. In today's virtual computing world, "tin" means little. Even a small telco will eat Akamai's global infrastructure for lunch in terms of "tons of tin," so I am not impressed by these numbers in the slightest. Its what you do with the tin, and most importantly what you charge in the face of your direct competition, that really matters.
Likewise, the cloud is not an infrastructure game, either. It is an economic game. Companies looking to replace hosted tin with cloud tin must have an economic reason to do so. The pure "modernization" argument accounts for very little. They must have volatile compute demands.
But, and I return to this as the most important thing, a resilient strategy is the most important strategy. If you have a single-vendor strategy you lose many advantages of a multivendor one-both in terms of your availability of service and in terms of being able to steer your prices. CDN relationships rarely cost a lot unless you are using them, and where you are using them you are generally charging for them. The logic of "phasing out" a relationship with Limelight would seem to be an unnecessary divorce in a market where polygamy is not only permissible, but in fact is desirable and expected.
If that phase-out is true, my gut instinct tells me there is more to it than an announcement of an "exciting new deal with Akamai." Perhaps Rackspace sees the only way to catch up with Amazon is to partner with Akamai.
And frankly, having watched Amazon's resounding progress recently, perhaps the only way Akamai can see being able to catch up with Amazon in the cloud space is by getting very close to Rackspace.
Interestingly, Rackspace is a far less mature cloud than Amazon, and Limelight is a much more modern CDN than Akamai, and one that appears to be agnostic in its service provision.
In some ways, I wonder if the long-term effect on share prices could be the reverse of what I noticed last Friday.
Octoshape's global delivery and video optimization technologies will help Akamai deal with large network traffic increases in the future.