I just realized that in my previous post about attending HPE’s Discover Conference as a speaker I said I would post the presentation after the event. I am a bit late, but here it is.
Good morning and thank you for joining us for the VDI on HPE Synergy session here during Discover 2019. My name is Rich Faulkner, and I’ll be leading our informal discussion today. A little about me. I have been involved with the End User Computing community for over 20 years. I started my career as a nuclear engineer on submarines in the Navy. In fact I have a picture of myself in a Hawaiian shirt at the North Pole. When I exited the service I decided to re-career myself in the IT field. I am currently a Solutions Architect at SageNet. I am certified in Server and Workstation virtualization. And recently I was awarded the title of Citrix Technology Professional. You can find me on LinkedIn and Twitter. Please let me know if you have any questions as we go through the discussion.
Today we will be discussing a deployment for a customer of mine. They are a large law firm in Oklahoma that has been in business since 1952. Currently they have over 250 employees with over 175 lawyers and supporting staff. They represent many of the large energy companies in Oklahoma. In addition, all aircraft titles in the US are handled by their firm. And my favorite thing they’ve done is negotiating the purchase and move of the Thunder to Oklahoma City. I’d like to say they consulted with me on that, but I’d be lying.
We engaged with the customer initially because they were planning to deploy a new VDI infrastructure. They had received several quotes and were unsure of the best way to proceed. They had been utilizing Citrix XenApp for many years and wanted to extend virtual desktops to their end users to add more flexibility in the way they were able to work. They had already configured their desktop image and run some testing on their current environment and applications. Unfortunately, their current infrastructure was aging. They currently had two VMware clusters. The first handled all their server virtualization needs while the second provided the resources necessary for their Published Applications being provided by Citrix and Application virtualization. They were utilizing HPE DL 380 Gen 8 rack mount servers. These servers were purchased at different times, so the clusters had differing models of CPU and levels memory on per host, making the infrastructure decidedly inflexible. They also had two additional older servers which they used for testing. Testing was limited because of the hardware capability, and because these even older servers that were used for could not accurately reflect their actual environment. With the upcoming VDI deployment, they were going to need more hardware, and they did not want to add a third cluster with another hardware type. They were exploring both rack mount and blade infrastructures.
By engaging with us as a partner they were able to determine the best solution to move forward. We followed out standard consulting methodology of Define, Assess, Design, Deploy, and Monitor. First, we worked with the customer to define the business case for Desktop Virtualization by creating a high-level roadmap and prioritizing the activities. We rated the key business drivers so we could prioritize the work correctly. We also reviewed their current environment to identify potential problems. We performed an assessment of their current infrastructure to determine where they were at now from a utilization and capacity standpoint. Utilizing their existing ControlUp monitoring package we were able to get a good view into the performance of their environment both live and historically. We then evaluated their VDI expectations to determine the required hardware. Using our proven consulting methodology we were able to determine the different types of users they would need to support this. We then determined the architecture required to satisfy the key business drivers and determine the success criteria. We determined the User Groups and identified their needs for Mobility, Security, Personalization, and Workload. Then then assigned VDI models to each of the groups. Then, we defined the applications that would be needed and where it would be best to install them (desktop image or published application). Then we could proceed to the design phase. We designed the architecture necessary to support their current workloads. We then calculated the hardware needed for each user type’s virtual desktop, and then calculated the required server specifications. We also calculated some expected growth workloads, and then the VDI workload. This permitted us to set the minimum specifications required for the replacement infrastructure. They had already received some hardware quotes, but it was determined the hardware resources specified would not meet their needs. In fact, one quote was for fewer resources than they were currently utilizing. The customer evaluated proposals for HPE Synergy, Cisco UCS, and Dell FX and both HPE and Dell rack mount servers. They wished to be able to move to a more condensed infrastructure and away from rack mounted servers, but they wanted to evaluate each option.
They decided to replace their current infrastructure with HPE Synergy frames and 12 Compute modules. Synergy would permit them to have a uniform hardware platform that was flexible enough to add more capacity in as needed. By doing so, they could compose the clusters and flex the capacity up and down as needed for their server and desktop virtualization structure and even have spares that they could use for testing and development. The environment was faster of course with the newer processors, but they also saw improvements in decreased network congestion and noted improvements easily seen in their ControlUp Monitoring Package. Nimble storage provided the space and performance required by the virtual desktop environment and the required I/O performance . In addition, with the frames, once they migrate all their workloads over they will remove the Fiber switches and will be able to connect the storage directly to the Synergy frames. The end result will reduce their racks from 4 down to 2. One View in the composers gives them the ability to apply profiles as needed and easily move resources around between clusters. InfoSight gives them the data they need for both the Synergy and the Nimble SAN for support and predictive analytics. They are also looking forward to more innovative features like the Nimble SAN being able to be configured in One View. Future capabilities permit them to use photonics as they become available and they are already evaluating Nvidia graphics cards either as a mezzanine add on or adding graphics modules in to the synergy frames themselves. They are very pleased with their decision.
If you have any questions about this, please feel free to comment on this post.