MeriTalk recently connected with ViON’s David Kushner, senior vice president of sales, and Chip George, vice president of public sector at Nutanix, to talk tech during the pandemic and beyond. From the evolution of virtual desktop infrastructure (VDI) to agencies’ demand for hybrid cloud, Kushner and George assess how government needs are changing and offer advice for agency IT executives looking ahead to 2022.
MeriTalk: What challenges will IT leaders face as agencies begin to bring employees back to the office? What will stay the same, and what will change?
Kushner: I don’t know if we’re going to see people 100 percent back this year. I believe we will see a hybrid, maybe with folks going in one-to-two days a week. I think we’re going to see more of the same VDI requirements and more remote work, potentially even an increase to accommodate workers moving out of the Washington metropolitan area and other city centers.
We are going to see agencies continue to move out of their data centers. They are either going to the cloud or a colocation space. They are going to require a hybrid multi-cloud environment with an “easy button” for maintaining and procuring infrastructure, particularly for organizations focusing on artificial intelligence (AI).
George: No one’s going to return to work fully the way they were before. Our third Enterprise Cloud Index found that 71 percent of respondents indicated their organization did not support any remote work prior to the pandemic. That number has plummeted to 11 percent today, and only 4 percent said they plan to go back to their pre-pandemic approach.
So if at least 85 percent are still going to work from home, that has to be optimized to reduce security risk. That’s accomplished with a VDI capability that gives the end user a standard image that is managed and controlled by the IT team with a back end infrastructure that can scale to users’ needs.
MeriTalk: How has VDI affected how agencies do business, and how do you see VDI deployments changing over the next year?
George: I think you’re going to see more as-a-service approaches to VDI. Some agencies are employing traditional VDI models that are run by a set of software on premises and the images are accessed remotely. That’s appropriate in some cases. Desktop-as-a-Service (DaaS) has gotten popular for the folks that are very cloud forward.
Kushner: Typically, 80 percent of IT budget goes to operations and maintenance, and 20 percent goes to new CapEx requirements. With as-a-service, we try to flip that and allow the organization to upgrade its infrastructure and scale as slow or as fast as they need to by utilizing a different color of money. They can use our cloud management platform to not only remotely shop for the hardware and software they need to spin up instances in the cloud, but also to integrate their applications, manage the governance of the as-a-service model, and see their spend on a single pane of glass.
MeriTalk: Some agencies implemented VDI for the first time because of the pandemic. What unique challenges did they face?
George: If VDI was brand new in their organization, in most cases they needed to upgrade their back end infrastructure into a true private cloud. But first, because of the urgent need to get people working remotely, they spent a lot of money to make their circuits bigger. Afterward, they made improvements to the back end. Some went to a true private cloud – implementing service-level agreements for service delivery and chargeback for services.
MeriTalk: How has the perception of VDI evolved?
Kushner: We’ve gotten so much better at understanding VDI requirements. The technology is easier to use. As long as you do the right planning before implementation, you’re going to have success.
George: VDI is more performant, and end users like it. In fact, government agencies were some of the first to deploy our DaaS, Frame, which is designed to support the most robust use cases, like computer-aided design and manufacturing. Also, the per-user cost model gives executives a predictable and straightforward way to manage their spend.
MeriTalk: What are some of the challenges customers face with deploying VDI to multiple locations?
George: Security is a challenge. If you’re deploying VDI through different VPN connections in 10 places around the country, you have to monitor and collect data on all of those connections. That means you need more robust infrastructure security.
If you’re running VDI for thousands of people, you are collecting thousands of sets of data, and your security applications need to be able to keep up with the volume.
As the number of locations increases, infrastructure increases in complexity. That can be remedied by a solution that converges all of the typical functions of a data center – servers, storage, networking, and a virtualization layer of software – under one operating system.
Kushner: If the technology infrastructure is easy to use, easy to manage and maintain, and easy to add to when a flex in data and capacity requirements is needed, the IT management burden shifts away from VDI and the hyperconverged platform and allows leaders to focus on cyber and data management challenges.
MeriTalk: What trends do you see taking hold in the data center in 2021 and in 2022?
George: The primary trend that is taking hold in the data center is hybrid cloud. In our Enterprise Cloud Index, 87 percent of Federal respondents said hybrid cloud is their ideal operating model. Their plans call for more than doubling their hybrid cloud use within one year, and to about 74 percent penetration within five years.
Kushner: Beyond the data management and the cybersecurity issues we’ve discussed, AI and machine learning (ML) are becoming a lot more mainstream. Also, learning how to better manage all of the infrastructure, whether you’re running applications in the cloud behind a firewall or in colocation is really important.
Lastly, aligning cost to budget for critical mission objectives is going to be extremely important. I’m excited about the as-a-service model and the role it can play in helping agencies modernize IT infrastructure and what that means to our customers.
MeriTalk: As agencies do more with AI and ML, what can they do to alleviate some of the stress on IT infrastructure?
Kushner: Agencies should focus on one or two specific use cases. The as-a-service model will enable them to scale as slowly or as fast as they want to go. That way, agencies aren’t over-buying or under-buying capacity, and they can ensure the projects align with their mission requirements.
George: AI and ML applications can be unpredictable in their performance and demands. They need access to the right data to make decisions. A lot of times that means they need to be exposed to more and more data. If we can optimize how that’s done, we can do these new and exciting things better, whether it’s cyber analytics improving agency processes. But you can’t do all that if you don’t have a flexible infrastructure that can scale.