In my last post, I laid out the case for and against various approaches to cloud-hosted applications. These topics were all discussed at a recent industry conference on cloud application platforms. Software as a service (SaaS), platform as a service (PaaS), and infrastructure as a service (IaaS) were all deemed to have associated benefits and drawbacks.
The question posed to me several times at the conference was this: why does Citrix care about cloud applications per se? XenApp (terminal services) seems like old school technology. Even XenDesktop (virtualized desktops) appears targeted at legacy applications from the pre-cloud era. More direct cloud offerings such as XenServer and OpenCloud seem more like infrastructure than application platforms. Thus, attendees at the conference could not initially understand what role Citrix might have to play.
Judging from the reactions I received, I think many were surprised at how obvious some of the answers were. Here is a more detailed version of what I said.
Infrastructure as a Service (IaaS)
In most ways, this scenario is the simplest to explain. Simply moving existing IT assets to an off-premise infrastructure does not materially change anything from an application access perspective. Just as XenApp might host an application on-premise today to provide remote access, backward compatibility, or for just about any other reason, XenApp can and should provide that same service tomorrow from the off-premise cloud.
Furthermore, since the change to public cloud is piecemeal, everyone agreed that so-called “hybrid” clouds that straddle on- and off-premise will be a reality for many years to come. In turn, network performance remains the thorniest problem traversing clouds. So, where an application may be directly installed on clients today, there may be new needs to host applications off-premise on XenApp or XenDesktop to minimize the network demands and maximize performance.
The role of Citrix actually goes beyond even what I said that day. Our networking products such as NetScaler VPX can be used to accelerate network performance in a variety of ways. Including our support for virtual networking components and the fact that many public clouds will be run on XenServer, the IaaS world will not be safe from Citrix.
Platform as a Service (PaaS)
Ah-ha, many responded, but we are porting these very applications to the HTML web. In fact, as we move to the cloud, we are planning to port applications to these application platforms like Azure and vFabric. I agreed, this does change the IT landscape in some ways. However, there are number of issues lurking below the surface.
We had all agreed that any shift to a public cloud is piecemeal in nature. The general notion is to move a data set and all associated applications at the same time, but not all applications and all data at once. Further, a generally accepted statement was that the average lifetime of an enterprise application is about a decade. Confronted with that statistic, many attendees admitted to concerns over the return on porting all existing applications to the web or a cloud application platform. Further, some important ISV applications out of their control may never be ported.
Enter Citrix. Using our infrastructure in the cloud, it becomes possible to move almost any application set to an off-premise cloud. IT can port any applications that provide sufficient return, but they can host the rest on Citrix infrastucture. Viewed this way, Citrix is actually an essential component to any cloud strategy based in the realities of cloud economics and technology.
In fact, for both the IaaS and PaaS cases, Citrix technology allows IT to leverage cloud infrastructure while simulataneously delivering applications to virtually any device. The recently released Project GoldenGate from Citrix Labs shows how Citrix can streamline application delivery and user experience for mobile devices as well as more traditional computing form factors.
Software as a Service (SaaS)
Many attendees thought this was the simplest part of the puzzle. As the vendor of GoToMyPC and GoToMeeting, obviously Citrix would think about moving those SaaS products to a cloud platform. I readily conceded someday that decision might be made by others well above my pay grade, but that was not a key motivator for my attendance.
Rather, I pursued a different line of reasoning. When asked, most agreed that any unification of user-facing IT processes materially reduces support costs. The more pervasive the solution, the more savings achieved.
I pointed out that in most respects the traditional Citrix business of application and desktop virtualization very closely resembles an IT hosted “software as a service”. Building out our offerings to support a single point of access, and therefore a single point of control, that included third party SaaS offerings makes a lot of sense. In fact, it makes sense for many PaaS-style applications, too.
Given the ubiquity of our client technology across devices, the goal of unifying access to applications is actually quite conceivable. Whether the application is subsequently delivered to a client-based web browser, streamed to a desktop, or delivered as a hosted screen to a consumer-owned device to protect sensitive data, it all works the same way to the end user.
This concept produced a less certain response. While no one outright rejected the idea, I don’t think they were entirely convinced…yet. With the lack of objection, I chalked this up to a healthy “show me” outlook.
I could actually write several more paragraphs on the various cloud scenarios in which Citrix technology makes sense. Even so, I believe it should be apparent that any enterprise looking to embrace cloud applications and off-premise hosting needs to include Citrix technology in their plans. The XenApp offerings alone provide a proven solution to several thorny problems presented by cloud computing paradigms.
For me, the real questions are whether and how many decades it will take for the industry to evolve to a point that hosted applications and remote access to them becomes irrelevant. My guess is that my son is more likely to encounter that world than I am.
Agree? Disagree? Feel free to leave comments and continue the discussion!