This week I had the chance to talk to Ali Din, CMO at dinCloud. dinCloud featured prominently in our early exposure to DaaS, and we had many conversations with Ali and CTO Mike Chase leading up to the publication of our 2014 book Desktops as a Service. I thought it would be cool to circle back and check in on them to get their opinion on how DaaS is progressing, both at dinCloud and with their customers.
In general, dinCloud has seen the same maturation of use cases as everyone else. Customers are coming to them with specific use cases in mind, which has changed the way the DaaS product lines are shaped. Looking at dinCloud’s product line today reveals a large difference from what it looked like back in 2013, where they basically sold a single product. Today, dinCloud offers targeted Hosted Workspaces to support different use cases:
dinHVD for the most flexible “VDI from the cloud” type experience. dinHVD deflivers a full persistent or non-persistent desktop. It supports any endpoint device, multifactor authentication, local USB devices, and multiple monitors.
dinRDS is similar, but as you can extract from the name it delivers session-based desktops and applications. It also supports multiple monitors, multifactor authentication, any endpoint device, and local USB devices.
dinApp is a product specifically designed for applications. It supports running different versions of the same app (a use case that they’ve come across). It supports the same features as dinHVD and dinRDS, while allowing you to deliver legacy applications that may not work on newer platforms.
dinDaaS seems similar to dinHVD on the surface, delivering a full desktop, but it’s been created with security as the primary focus. The only client for dinDaaS is a Chrome plugin that works on Chromebooks or the Chrome browser (Mac, Windows, and Android). Beyond that, the other primary limitation is the lack of multi monitor support, which makes sense since it runs in a browser window. Local USB is restricted and configurable, as is copy & paste. Local printing is not possible, but network printers and printers supported by Google Print are possible to use.
One of the main challenges I come across relates to application data. Specifically, if I have an application with a client on my desktop that needs data that lives in the datacenter, moving that application to a desktop at a DaaS provider presents a tough problem to solve. I asked Ali about this, and it turns out this is something they recognized early on. They call it the Pure Physics Problem (wherein you simply get better performance the closer you are to the data), and because of it they realized that in order to be a complete DaaS platform, they also needed to offer IaaS as well.
To be clear, the IaaS that they offer is not the all-reaching IaaS that you see from other cloud platforms. Rather, it’s IaaS as it pertains to DaaS. In the case of application data, most of their customers move it along with their desktops when they begin to use dinCloud. This is a fine solution, but it brings to light something that isn’t always obvious:
When determining the cost of DaaS, you also have to factor in the cost of running your application backends at the same provider.
Of course, there are situations where you can’t move the data from the datacenter (for example, if you have a mix of DaaS and on-prem desktops). In those situations, dinCloud will work with you to set up a VPN that connects your isolated network from within dinCloud to your on-prem network. The isolated network on the dinCloud side is a software-defined network that the customer controls.
Another challenge that often comes up is Active Directory. Though it’s not insurmountable, deciding how (and to what) your users will log in is something that needs to be addressed ahead of time. dinCloud customers use a mix of strategies that range from federation and synchronization to remote domain controllers on the customer’s domain. In some cases, dinCloud provides nearly all of the datacenter-based services, in which case one AD running on the customer’s dinCloud system takes care of everything.
Why are their customers using DaaS?
I keep looking for new use cases for DaaS, but to be honest I haven’t seen one for a while. The primary reasons dinCloud sees for DaaS adoption are basically the same as we’ve heard for years. They have customers that just want the system up and running as fast as possible, and dinCloud has put a lot of resources into the onboarding and management to make that possible. Others don’t have enough IT resources on staff to build, deliver, and maintain an environment on their own, so they choose DaaS to take some of the load off.
Of course, predictable costs are a big benefit, and dinCloud has dedicated themselves to fixed costs rather than usage-based (and therefore erratic) pricing used by other providers. It’s not that they’re charging more for everybody to make it a flat rate, it’s just that they’ve built their pricing with the erratic usage already in mind.
dinCloud has been around since 2011, which basically makes them the geezers of the DaaS space. Together with Quest Systems, NaviSite, and some old-guard Desktone people left at VMware, they can kick back and talk about the “good ole days.” Really what it means, though, is that dinCloud has been around long enough to weather the storm. They’ve seen DaaS move from the periphery of IT to the forefront, which is important when one of the primary reasons for not delivering DaaS is “trust in the provider.”
Overall, the conversation I had with Ali further reinforces my positive outlook on Desktops as a Service. We're not seeing a huge number of new use cases, but we are seeing more mature product offerings that are solutions to problems rather than solutions looking for problems. It's not going to happen fast, but DaaS adoption will keep chugging along.