Yesterday Citrix announced that they’re buying a company called NetScaler for $300M. NetScaler is a hardware vendor that sells a range of networking-type equipment that does application load balancing, caching, compression, and acceleration. (Think F5 or Cisco, but with more of a focus on the application data moving over the wire instead of the networking data.) NetScaler is used in big companies that have to deal with major web application traffic, including Amazon.com, Google, and MSN.
Rather than repeating the news stories that are starting to emerge about this deal in just about every tech magazine, I’m going to explore what I think this deal means for the future of Citrix.
The Fork in Citrix’s Road
I think Citrix came to a critical decision point a few years ago. They had to decide what kind of company they wanted to be and what their long-term vision would be. The problem was that Citrix grew up as an “add on” product for Microsoft Windows. This was a risk for two reasons:
- There was always the chance that Microsoft would just incorporate the Citrix functionality into the “base” Windows / Terminal Server product.
- As the world moves forward, the line between server-based computing and web computing will blur, and Citrix’s “niche” will be assimilated into the maze of mainstream application architecture.
To mitigate these two risks, Citrix had two choices. They could either:
- Become an application management framework company and broaden their horizons to focus on both “server-based” and “traditional” applications. In a sense Citrix could manage all applications no matter where they were—remote Terminal Servers, local installs, etc. This is the direction that server-based computing software vendor Jetro is taking and is probably pretty close to what companies like Altiris or Scriptlogic will look like in a few years.
- Alternately, Citrix could become more of an infrastructure company and focus on the hardware and software needed to deliver applications (web, Windows, and otherwise) to users.
Obviously Citrix chose the second choice.
How does this fit into the future?
If there’s one thing that apparent about the future of IT, it’s that IT infrastructures are becoming more modularized and service-oriented. (That is to say that many individual IT components will work independently and provide “services” to each other.) For example, application execution will be a “service” that’s provided by back-end execution servers (like Terminal Servers or IIS Web Servers today, although these technologies will converge eventually). Application user interface presentation will be a “service” that’s provided by the client. (This is provided today by client rendering engines in web browsers and ICA or RDP client software.)
The future will be a services-oriented modular environment, with application execution services and presentation services being just two pieces of a bigger service fabric.
Microsoft already owns the application execution services layer in our world (Windows, Terminal Server, and IIS), and its clear that over the next three to five years they will continue to evolve in the presentation services area by adding more presentation services capabilities (Bear Paw, Avalon, etc.). This means that eventually Citrix won’t be able to provide as much value in the presentation layer as they do today with Presentation Server.
So as Citrix first outlined in 2003, they will focus on the “access infrastructure.” In my mind this means linking together users, client devices, and applications. (Again, remember that “applications” in this case is a combination of execution services and presentation services, and that “applications” can be web, Win32, or .NET-interconnected.)
Last year’s “access infrastructure” from Citrix was purely at the presentation services-layer. (Of course they had some portal products and stuff but no one cared since they didn’t easily fit into this whole services-oriented application model.) This year Citrix added some capability at the security infrastructure layer with their Access Gateway SSL VPN appliance. Yesterday’s announcement about the acquisition of NetScaler will allow Citrix to continue to add capabilities at the layer where the application meets the network. In a sense Citrix is inserting competence in between the application execution service layer and the application user interface presentation layer. All of this, though, still falls within the grand realm of linking together users, client devices, and applications. (While Citrix calls this “access infrastructure,” NetScaler calls it “application delivery services,” which is probably more technically accurate.)
The Reality for Citrix
So what does this NetScaler acquisition actually mean for Citrix (besides an extra $60M in revenue next year)?
First of all, if there was any doubt in anyone’s mind that Citrix wouldn’t compete with the Cisco’s and Nortel’s of the world, that doubt can be erased now. Switches and routers have become commodities. (That’s why Dell is selling them.) Today’s high value networking equipment is the application-focused (OSI Layer 7) intelligent infrastructure hardware (VPNs, load balancers, accelerators, etc.) like what NetScaler offers.
So for now this means that Citrix can offer capabilities at the presentation, security, and delivery layers (with a long-term shift towards delivery).
In some ways this kind of makes Citrix big and goopy like Sun. Sun has hardware, software, presentation services, security and identity services, infrastructure services, portals, storage, etc. Citrix’s wide range of capabilities means that Citrix is really going to have zero-in on their focus here. Will they force people to buy their software to get the best value from their hardware? Will they force people to buy their hardware to get the best value from their software? Will they have hardware and software that is tangentially related but often bought by different customers?
One of the interesting footnotes of this acquisition (like the others of Citrix) is that in the press release, the NetScaler executives mentioned that they were excited to get access to Citrix’s strong reseller channel. The challenge is that Citrix’s reseller channel is primarily focused on Windows server-based computing solutions. While in many ways that’s not that different than an application delivery focus, it will be tough for the “standard” Citrix partner to make the philosophical transition to be able to sell what NetScaler offers.
Then again, we may see the current Citrix channel take interest in the “traditional” capabilities of the lower-end NetScaler devices (load-balancing, packet shaping, etc.) while the high-end capabilities are left to channel members who focus on high-end application consulting.
At the personal level, Citrix professionals will need to make a decision about which direction they want to take their careers. Will they focus on infrastructure services or presentation services? Each must decide because both are too much for one person to handle. (In case it’s not obvious, I personally will focus on the presentation services—the software layer that links user interfaces to application execution. I will not focus as much on the access infrastructure path that the presentation services layer will ride on.)