by Rick Dehlinger, Citrix
Hello again my community friends!
As many of us are skittering about, closing down 2016 and preparing for the Holidays, the first team of (Elven?) geeks has kicked off a wave of collaborative projects that should bring good tidings and cheer to our community. It’s an amazing group with a diverse set of talents, and I’m both excited and honored to get a chance to work with them all. None of these community members are new to Citrix (as you might expect) and they all share in their desire to give back to our community:
- Dave Brett (@dbretty, https://uk.linkedin.com/in/dbretty, http://bretty.me.uk/)
- Eric Haaverstein (@xenappblog, http://xenappblog.com/about/)
- Esther Barthel (@virtues_it, https://www.linkedin.com/in/ebarthel, http://www.virtues.it/)
- Jason Policy (https://www.linkedin.com/in/jason-policy-28ab4416)
- Jason Samuel (@_jasonsamuel, https://www.linkedin.com/in/jasonsamuel, http://www.jasonsamuel.com/)
- Joel Piper (@ctxpipes, https://www.linkedin.com/in/piperjoel)
- Ryan Snedigar (@ryansnedigar, https://www.linkedin.com/in/ryansnedigar)
Before I get too far into the actual projects and our execution plan, I need to give a quick shout out to Cisco. Unless you were on a blessedly disconnected sabbatical (or were living under a digital rock), you probably caught that earlier this year Cisco released a family of Hyper Converged Infrastructure appliances (HCI) called HyperFlex. It’s some pretty bad a$$ed gear, and we’re fortunate to get plenty of hands-on time with it in the Silverton lab! HyperFlex is based on Cisco’s UCS C-series servers, and leverages their UCS Fabric Interconnects for network connectivity and management. We’ll be sharing our hands-on experiences with the joint solution through a variety of projects in the Silverton Hybrid Cloud Lab, but if you’d like to learn more about Citrix on HyperFlex in the interim, check out http://www.accelerateyourworkforce.com/.
Speaking of projects… now that I’ve revealed the techies involved and the gear we’re running this first wave on, let me introduce them! I encouraged the team to apply the basics of the scientific method as they formulated their project plans. I’ll let the details of their projects unfold over time, but to get the ball rolling I’ll share their starting points – the questions they’re trying to answer.
The first project I’m calling “Hands on with HyperFlex.” In many regards, HCI is the new, shiny thing in datacenter tech. It promises to break down the typical silos in IT organizations, and bring a whole new level of simplicity and ease of management to customers of all shapes and sizes. Citrix VDI is one of the primary workloads people are running on HCI, and for good reason – these appliances usually remove one of the biggest obstacles to successful deployments: storage performance. Led by Joel Piper, this project will explore the following questions:
- How long will it take to get Cisco HyperFlex up and running?
- What are the necessary steps to be completed in order to have HyperFlex prepped for a XenApp/XenDesktop installation? How easy (or difficult) is it to complete the process?
- What resources – guides, how-to articles, and documentation – are available to assist the average customer in this endeavor, especially a customer who has not been formally trained as a datacenter engineer?
- How well suited is HyperFlex to running MCS-provisioned workloads? How quickly can catalogs be created, updated, or rolled back?
The second project I’m calling “Hands on with Workspace Environment Manager.” Earlier this year Citrix acquired Norskale, a company built by friend, fellow CTP Alumni, and Citrix rock star Pierre Marmignon (@pmarmignon). Norskale’s products have since been re-branded as Citrix Workspace Environment Manager, and are now available to Citrix customers as part of Enterprise and Platinum editions of XenApp/XenDesktop. Besides powerful environment management and workspace configuration capabilities, WEM includes some fantastic optimization technologies that promise to substantially improve the performance, user experience, and density of XenApp/XenDesktop workloads. Ryan Snedigar and Eric Haavarstein are teaming up on this one, and they’re exploring the following questions:
- How does WEM impact the performance, user experience, and density of XenApp/XenDesktop workloads?
- How does HyperFlex perform under load from modern Operating Systems like Server 2016 utilizing:
- …only policy enhancements from MS and Citrix with PVS?
- …only policy enhancements from MS and Citrix with MCS?
- …WEM (Norskale) alongside policy enhancements from MS and Citrix with PVS/MCS?
- Has the HyperFlex provided an improved level of enablement, optimization, or other tangible advantages for the IT administrator?
The third project I’ll call “Virtualizing Skype for Business.” Almost everyone operating in the desktop virtualization space has been touched by Skype for Business in one form or another, and Citrix gives people the best virtualized S4B experience out there. Recognizing S4B’s prevalence in the market, Jason Policy’s project will dig into this popular business app and seek to answer some of the questions you might be asking if you’re about to deploy this popular application on XenApp/XenDesktop. He’s starting with the following questions, and it will be interesting to see what others he uncovers on his journey:
- What’s the relative impact on performance and scalability when running Skype for Business via XenApp or XenDesktop?
- Does that change when some of the users leverage the multimedia features?
- What impact do the Citrix RTME optimizations have on scalability and user experience?
The fourth project I’ll call “NetScaler VPX on HyperFlex.” On the surface this one might seem like a head scratcher. After all, pretty much every modern XenApp/XenDesktop deployment includes NetScaler, regardless of whether VPX’s or physical appliances are in use. For customers deploying HCI, however, I’m hearing a slew of other questions come up. Upon further introspection, many of these questions are pretty universal, so we jumped at this project. Esther Barthel and I are working on this one, and plan to explore the following:
- Is running NetScaler VPX on an HCI appliance OK?
- If I run it alongside my desktop/application workload, what does it do to the scalability/user experience?
- How ‘big’ should these appliances be?
- How should they be configured for satisfactory results?
- What other things should I be concerned about if I deploy this way?
The fifth project I’ll tease at, then leave Jason Samuel to spill more beans over time. One of the cool ‘toys’ I’ve got in the lab is a NetScaler SDX 14100 appliance. This is the first time I’ve had access to one, and maaaan – it’s sweet! Jason’s been doing some really cool work with NetScaler SDX recently, and has joined us in the Silverton lab to polish up a couple of articles that I think you’ll enjoy.
The final project is a quick but fun one. Citrix Summit (our worldwide partner event) is coming up in a few short weeks. Google will be there in force this year, and I got tapped to work with them to put together an instructor-led training course focused on Citrix on Chromebook. This is a pretty cool use case for our recently released Federated Authentication Services technology, which, when paired with the right configuration of the front-end NetScaler, provides a great single-sign-on experience for users on these flexible, low cost devices. Fortunately for me, Dave Brett has already done this sort of advanced NetScaler config as it’s well beyond my normal area of expertise! For this project, we’re working together to build the demo environment to support the ILT at Summit, and Dave’s going to write up and post an article on the complicated parts, allowing the training participants (and the rest of us!) to learn how it’s done and done right.
I’ll wrap up this article by sharing a tidbit of how we’re planning to do all this in parallel. There’s some ambitious work here for sure, but by working together we should be able to really amplify our learnings and increase the quality/quantity of what we share back to the community.
These projects will be run on two separate, four node clusters of Cisco HyperFlex. We’ve built a common baseline environment we’ll be sharing – Lab60. All of our supporting infrastructure (domain controllers, SQL servers, XDCs, FAS servers, file servers, WDS, LoginVSI datastore/launchers, etc.) are running on one cluster, and we’re running and measuring the different workloads from our projects on the second. We’re working together to establish the baseline workload(s), to which we’ll compare the output from the various tests we’ll be doing over time.
One important and exciting note: the Silverton HCL leverages Eric Haavarstein’s automation framework! This is making the creation/deployment of the different workloads and environments MUCH easier. It’ll also seriously simplify the scale/performance testing work we’re doing, allowing us to get a lot more done and ensure consistency in the results. You’ll get to hear more about this framework over time, and I think you’ll be BLOWN AWAY at what it’s capable of. Suffice it to say, I think every Citrix environment needs it – and Eric makes it available for a song. If you can’t wait to hear more, head on over to http://xenappblog.com/ for more information.
That’s all I’ve got for now. I’ll close this article (my last of 2016) by wishing you all a Happy Holiday – I look forward to working with many of you in the coming year as we take hybrid cloud with Citrix to a whole new level. Until then, cheers!