When the coronavirus pandemic forced everyone out of offices, IT teams scrambled to obtain the necessary hardware and additional software licenses necessary to support entirely remote operations. At the same time, they had to match users’ present and future needs with the appropriate infrastructure. That, combined with many other unknowns, compounded the stress and urgency to keep business running smoothly around the world.
If we’re being honest, many organizations struggled to adapt to the recent shift to remote work because they simply weren’t prepared for such a major upheaval. Duane Morris was not one of those companies.
Duane Morris LLP is a law firm founded in Philadelphia in 1904, and more than 100 years later, the firm employs more than 800 attorneys in offices across the United States, Europe, and Asia. I’ve been the CIO since joining the firm almost 20 years ago. I’m responsible for all the enterprise technology, including data, network admin, security, training, help desk, development, videoconferencing, and telecommunications, among other things.
About eight months ago, my team and I began discussing ways to upgrade our existing infrastructure. We wanted to modernize it and enhance our compute capabilities by expanding our current hyperconverged platform to enable more remote work. Little did we know how important that would become in the months ahead.
Simplifying and Optimizing Our Multi-Vendor Partnerships
We’ve been using Citrix for more than 20 years. But three years ago, as part of our negotiation with VMware to renew our enterprise agreement, we included enterprise-wide Horizon View licenses. We considered using virtual desktop infrastructure (VDI) as an enterprise platform for our desktops, but made the decision not to go that route. However, it really stood out as a robust remote access platform.
We ended up using both Citrix and Horizon View as our remote access platforms. Two remote access platforms give people a choice and also create a level of redundancy—and it was a complete lifesaver when we went into lockdown. We basically had a Horizon View license for everyone across the firm.
We next moved to our data center storage options. Our data center has always been primarily Hewlett Packard Enterprise (HPE, and 90% of our servers are currently HPE. Our relationship with HPE goes all the way back to the EVA platform; we currently have the HPE 3PAR platform and the new HPE Primera, which we just acquired through the HPE GreenLake program.
HPE GreenLake is an ultra-scalable cloud-like experience for all applications and data wherever they reside, competing against the public cloud and traditional in-house purchasing models. With HPE GreenLake, you don’t have that initial large capital outlay. Instead, you’re paying for what you need, with extra capacity layered on top that you only pay for in the event you need it. The platform is fully managed by HPE, including updates, and at the end of the contract, you can upgrade that platform to the latest and greatest if desired.
We were in a position to either sign a new maintenance contract with HPE, consider purchasing new HPE equipment (Primera), or go with HPE GreenLake as an alternative to both or consider competing storage platforms. We moved ahead with HPE GreenLake for our primary storage solution, but we also liked the ability to add on to our current hyperconverged platform, HPE SimpliVity, which we were going to build out for enhanced remote compute capabilities. We needed more dedicated resources for compute when building out a VDI environment, and since we use both Citrix and VMware as remote access platforms, we established a plan to build out additional Citrix connections on the new HPE SimpliVity platform.
When I shared this model with our finance department, they were pretty happy with the way it calculated out over the next several years, compared to an upfront capital investment. There were significant savings versus buying it outright.
Adapting to the New Workplace Reality: Building Secure Virtual Systems Effectively
Many businesses put their plans on hold when global lockdowns began, but we knew this was the perfect time to move forward with our infrastructure and systems upgrades. This was partially due to the fact that we had already started the processes. We also knew that this setup was exactly what we would need to continue to work effectively, so it made sense technically and economically to continue moving in this direction.
Like everyone else, our staff’s remote access has surged over the past few months. Prior to the lockdown, 450 people was the maximum number of remote workers we’d ever had logged on at the same time. Within a day after all of our offices closed, that number jumped to more than 1,200. There’s no way we could’ve sustained that amount of remote access reliably with our previous infrastructure.
We’ve also experienced a huge increase in web conferencing, which we’re actively supporting. In fact, we recently hosted a firm-wide diversity and inclusion webinar that was attended by more than a thousand people. It was the largest web-based conference that we have ever hosted and it was also the most highly attended meeting that the firm has ever held, virtually or in person. We were really pushing the limits of our technology, not knowing exactly how it was going to work. The good news is that it worked very well.
Navigating and Negotiating Our Way to a Bespoke DRaaS Model
For disaster recovery, we’ve been using a co-location model for the last 15 years, which involved renting data center space from a provider while using our own equipment. You’re paying a lot of money for storage to sit there just so you can replicate data. We also needed to make sure that we had sufficient compute resources in the event that those resources were needed. So we were putting both storage and compute resources into the co-location with the hope that we’d never have to use it; it was just there for disaster recovery purposes.
I decided over a year ago that the co-location model didn’t really make sense for Duane Morris anymore. There were now smarter, more cost-effective ways to get the job done, and we began exploring disaster recovery as a service (DRaaS) offerings.
We put this on a slight hold to focus on a merger, and the timing led right into the pandemic lockdown. When our evaluation of solutions resumed, there was a new solution made available by Anexinet and HPE.
Once we’d chosen HPE GreenLake for our primary storage, this opened the door to explore Greenlake as our DRaaS. We liked the fact that with a simple change order, we could expand our existing HPE GreenLake capabilities with the DR offering without going through multiple contracts and agreements. In this model, HPE owned and managed everything, and there were other compelling points.
Typically, what drives down costs in the DRaaS model is the fact that you’re a co-tenant and are therefore sharing the costs. But you can’t be a shared tenant for all resources, because if you’re replicating, you need dedicated disk space to replicate. You also need dedicated resources for compute, because at least some initial setup needs to be there on standby for you to build out a VDI environment.
In the HPE GreenLake model, you can increase your pool or resources for what you need in the event of a disaster, and grow that compute environment as necessary, again, only paying for what you need, when you need it. This allowed us to have a sizable return on investment. On top of that, we would be using all new equipment that was totally dedicated to Duane Morris, so there was no shared tenant as was proposed with the public cloud option. This provided known resources if needed, in the event of a disaster. And everything was managed by HPE under the HPE GreenLake model, which provided a high level of security and the ability to utilize the resources in a production and/or test environment.
During this process, our HPE partner, Anexinet, was instrumental in working with HPE to create this option suited to our business. We’ve been doing business with Anexinet for about 10 years and thanks to their input, Duane Morris is the first customer using HPE GreenLake for disaster recovery.
My role hasn’t changed much over the past two decades, but the technology has changed significantly in that time. As we’ve upgraded, our footprint keeps shrinking considerably. We’ve already gone from several racks of storage down to one, with faster, more reliable, less expensive drives. That’s also why it makes sense to only pay for the space we’re using. It gives us added flexibility and reduces overall costs. In addition, our solution now is almost totally transparent, so it’s less effort to oversee storage and servers.
As data storage increasingly moves to the cloud, our data centers have evolving needs. And while you can certainly expand your resources and scale up, it’s also important that companies like ours are able to scale not only up, but also down, due to newly available cloud storage possibilities. HPE GreenLake gives us that flexibility.
Always Preparing for the Next Big Thing
While we always knew that we wanted an infrastructure with extra capacity, we couldn’t have predicted the sudden shift to remote work—we were just building resilient systems that would allow us to adapt quickly whenever the shift occurred. It never hurts to be strategic about your long-term plans. Circumstances like mergers, unexpected disasters, and a remote work paradigm shift can’t always be predicted, although you can always start preparing.
It doesn’t matter if your workers are in-office or remote. You can’t rest on your laurels when it comes to your infrastructure because it has to remain viable to support your people no matter where they are. From switches to compute to storage, everything has to remain current with an eye to what your employees are going to need in the future. This approach is why we were able to finalize our new infrastructure and systems in the middle of a pandemic that has incapacitated companies around the world. Others have lost out on a lot of time and money as they struggle to catch up.
Personally, I think the global percentage of office workers going remote (either part time or full time) is going to increase significantly across all industries. Our working reality has experienced a paradigm shift, forcing us to reevaluate how we all see the future of work. If you want your company to be able to compete in the months and years ahead, it’s not enough just to keep the lights on. You have to continue pressing forward and making investments now, before it’s too late.
The HPE GreenLake model provides us with a cost-effective approach to both scale up and scale down, as it relates to working in the office, working remotely, business continuity, and cloud. We also gained more flexibility and “hands-off” IT, which has allowed my team to focus on the needs of the business—which has been invaluable. This model is going to serve this new workplace reality brilliantly, and is the kind of adaptability and forward thinking that will allow the next generation of 100-year companies to emerge—and existing companies who can take advantage of these offerings to remain competitive.