Remote Desktop 2.0 Moving end user computing to the cloud!

So this is my tagline for my session at NIC CONF 2017 http://www.nicconf.com/remote-desktop-v20
But since its a large subject, I also decided to write about it as well. With the uprise of Cloud, it provides customers which a large opportunity in terms of choice when it comes to modernizing the way we can provide desktops / apps to end-users.

Now for many providing an application and desktop delivery solution has been tedious since you been needed to plan from scratch most of the time(CPU, Memory, DISK/IOPS, Capacity, GPU) and then you needed to assess the applications if you could deliver them from this type of solution, and of course you would need have the networking infrastructure in place. Decide on what kind of vendor you want to choose, what kind of endpoints you want to support and so on. Now of course after this has been setup you would need to maintain it, hardware, servers, update images, update the delivery software as well, and we also need a way to add in the mix of SaaS offerings which we want to our endusers to be able to access in an easy way.

End-user computing overview

image

As mentioned introducing the cloud has given us a lot of different advantages.

  • Rent capacity when needed
  • Pay for capacity only when needed
  • Unlimited Scale
  • Access to specific hardware without needed to invest
  • Dedicated resources
  • No need to maintain underlying hardware and hypervisor

Now these are just advantages aimed at making the infrastructure part a bit easier, but we still need to maintain the virtual machines, we still need to plan the setup of the software, remote access and it would still need maintaining. Now Cloud also introduces finished platform services, where we move higher up in the stack the software is fininshed setup and we also do some minor setup like adding an image with out LOB applications and we are given access to our applications.

Now with the introdution of all these different options, I decided to dig a bit deeper into each of them to provide you with an overview of all the vendors including features / strengths & weaknesses and where they position themselves in the cloud stack (IaaS/PaaS/SaaS)

image

So these are the vendors I’ve decided to focus on, in a kind of research document which I’ve labeled Remote Desktop 2.0 moving it to the cloud.

NOTE: If you have any other vendor which I’ve missed please contact me.

Now if we look at traditional vendors like Microsoft with RDs, Citrix with XenApp / XenDesktop delivering them in the cloud is supported in different forms.

Microsoft RDS: Can be setup in Amazon Web Services, Microsoft Azure and even Google Cloud platform, since all the pieces can be run inside Windows Server. Now RDS does not have any particular Cloud integrations, but it plays well with Azure. For instance the connection broker can now be integrated into Azure SQL to setup HA connection brokers, storage spaces direct is supported for RDS profile disks in Azure. I have blogged about setting up RDS Connection broker HA SQL database in Azure here –> http://msandbu.org/running-rds-2016-tp5-connection-broker-database-in-azure-ad/

Citrix XenDesktop/XenApp: Can be setup in Amazon Web Services, Microsoft Azure but not in Google Cloud platform still lacking NetScaler. Citrix however has provisioning capabilities built into the platform which allow them to provision resources in either AWS or Azure, they also have better capabilities in terms of integrations with Office365 with HDX optimization pack. I have blogged about setting up Azure Resources leveraging the built-in MCS provisioning engine in Azure here –> http://msandbu.org/delivering-xendesktop-from-microsoft-azure-using-azure-resource-manager/

Now looking part to this drawing, setting up RDS or XenApp/XenDesktop will provide us with a lot of control,

image

and still provide us with the same native capabilities as RDS and XenDesktop does, such as endpoint support, management capabilities, protocol support and so on, except that we cannot leverage the hypervisor connectors such as PVS in Citrix and RDS VDI which integrates with Hyper-V. Atleast we don’t need to worry about Storage / Compute and the virtualization layer.

Teradici Cloud Connect:
While not providing any full provisioning/management capabilities Teradici Cloud Connect is also supported with Azure leveraging for instance N-series with PCoIP or AWS Graphics Workspaces which truly shines with PCoIP Zero clients to leverage the GPU capabilities. I have blogged about Teradici Cloud Connect before and showing how it leverages the N-series in Azure  –> http://msandbu.org/test-run-of-teradici-cloud-access-software-on-azure-n-series/

Now for many these types of solutions are the safe options. They provide the customer with control, they are able to choose which type of resources and manage them like a regular virtual machine. They still have the advantage to scale-out as needed but they would still need to install agents, and can use the solution like they have used it before. The one issue I have is that none of these solutions have built-in mechanics to for instance scale-down automatically when not neeeded to leverage the “pay-for-what-you-use” model for public cloud. Also we are still stuck with the management pieces and components which we would need to be maintain as well.

So this first post, have been an introduction and a quick overview of some of the vendors, in the second post I will take a closer look at other vendors which have taken another approach .

You May Also Like

About the Author: Marius Sandbu

Leave a Reply

Your email address will not be published. Required fields are marked *