From the "imitation is the sincerest form of flattery" department, tuCloud released their free DaaS Engine this week (General Information PDF), which bears a striking resemblance to Citrix VDI-in-a-Box with its simple virtual appliance setup and grid architecture. I had a chance to catch up with Guise (Guy) Bule from tuCloud to hear what he had to say about it, and I learned that the tuCloud DaaS Engine actually serves three purposes:
- First, it's the backbone of tuCloud's own hosted desktop solution.
- Second, it's able to be licensed as a white-label solution for other hosted desktop providers.
- Third, despite the "DaaS" in the name, you're able to use it to host VDI desktops in your own company's environment. In fact, you can download a 25-user version for free.
When I asked Guy about the similarity to VDI-in-a-Box, he confirmed it was, indeed, the inspiration for the design of the DaaS Engine. If you recall, tuCloud was a huge proponent of VDI-in-a-Box's grid architecture before Kaviza was acquired by Citrix. After the acquisition, Guy says, Citrix started calling the VIAB customers and prospects to inform that it was an SMB only solution, which hurt tuCloud's business. Because of this, rather than use a solution that, in Guy's words, "Citrix bought to take out of the market," they built their own. If you've been paying close attention to the space, you might also remember that tuCloud partnered with Desktone early last year. The release of the DaaS Engine indicates that that relationship has fallen by the wayside.
The platform was solely created for tuCloud, but they're not ignoring other companies who want to use the platform to roll their own, private VDI solution, or companies that want to start their own hosting solution. The free version, which starts at 25 users, can be expanded based on participation in the forums. This is undoubtedly a mechanism to get real-world beta testers by rewarding them with free licenses. Participation in the forums nets an admin a certain number of points. There are ranks assigned to users at certain point values, and each rank entitles a user to an upgrade. Through upgrades, it's possible to work your way up to a multi-tenant, unlimited user license.
Of course, if you're not feeling like making 2,500 posts to get enough licenses for your company, you can buy into the product. The cost to do so is unclear, but signs point to it being relatively inexpensive compared to other solutions. From what I've seen, this solution looks decent, and I can see where some budget-conscious organizations would be more than willing to hop on board, and even contribute, for free licenses.
So what does the technology look like? For starters, this is a non-persistent desktop solution only. This shouldn't be a surprise for people familiar with Guy or with tuCloud in the past. Many an argument has transpired on twitter with regards to the merits of non-persistent versus persistent desktops, and tuCloud believes that non-persistence with some element of user environment management is the way to go. To them, the only thing that matters is providing the user with the things they need to do their job. Anything less and they're not happy. Anything more is overkill. To accomplish this, they use some of their own scripting to add some user state persistence features, but they are also working with Liquidware Labs on potentially integrated solutions.
The DaaS Engine ties into ESX currently, with plans to expand to Hyper-V and XenServer at some point (although I think they'd probably be best sticking with Hyper-V and ESX). It includes a secure gateway feature and a broker. Out of the box, it leverages RDP 6, 7, and 8, however it's unclear how RDP 8 UDP is handled through the secure gateway. Nevertheless, we have another example of RDP 8 being a top-notch protocol that people can leverage without having to feel like they're compromising on something.
Since this is the solution that tuCloud uses for their system, there is functionality built in for multi-tenancy and clustering, even between datacenters. Whether or not this is friendly enough to be used in an enterprise remains to be seen, but the few posts in the community forum indicate that people are having some success with it. According to Guy, it they have scaled the platform up to 5,000 users without issue, and they're currently working on a 20,000 seat deployment that has tested well in the lab but has yet to be rolled out.
The thing that hurts tuCloud with this whole deal is that there's a mixed message coming from them with regards to their focus. On one hand, there is talk about how this product is built for the sole purpose of running their service provider operation, and that their focus for the next twelve months will be in that area. That doesn't really change their agenda, though, since they've been playing in that space for quite some time and the competition is the same.
On the other hand, tuCloud has made competitors on two new fronts, and and they seem to have done so willingly. Since they can white-label it to other service providers, that puts them in direct competition with the likes of Desktone, who has also gone to that model in recent years (and whom tuCloud was once partnered with). Additionally, since they're acting as a vendor and making a product that people can use in-house, they're also competing with the likes of Citrix, VMware, Dell, Virtual Bridges, Microsoft, Ericom, and DesktopSites.
While tuCloud is quick to say that they're not trying to compete as a vendor, allowing people to stand it up in their own companies, especially for free, is certainly a competitive move. Plus, if they planned on keeping it to themselves, why go through the effort to make the install as simple as possible? If they want to enter that race, that's fine, but from a marketing perspective I think calling it a DaaS solution won't help them get into organizations looking to do their own VDI. The thought process a company has (or, at least, should have) is to first decide whether or not they want to implement a datacenter-hosted desktop solution. Only then do they decide whether or not the want to host it themselves or let someone else do it. If they look at the tuCloud DaaS Engine, they may not even consider it for their in-house solution, and may exclude it from the running altogether.
We can always use some extra hands stirring things up, but hopefully there will be more substance to what we see from tuCloud in the future compared to the past. Bravado has never been in short supply at tuCloud, known for lofty statements every so often, not to mention two changes in backend architecture in as many years. Hopefully we'll see some more action this time around. If tuCloud built on the VDI-in-a-Box model and really has made it more scalable for both service providers and in-house deployments, they could be on to something. They just have to settle on what, exactly, they're trying to accomplish, then follow through with all the energy we know they have.