
Your Infrastructure, Made Safe at the Edge
Because our world is increasingly driven by real-time applications and results, we have today’s rush to bring datacenter infrastructure to the edge, whether that edge is in a regional colocation facility or a store’s server closet. Commodity hardware will not suffice. Modern real-time workloads demand hyperscale datacenter-class servers and rack deployments, which means hyperscale infrastructure providers are uniquely suited to tailor their solutions into these new environments. Gartner notes that hyperscale cloud providers delivered and managed less than 1% of edge computing platforms in 2020. By 2023, that number will explode to 20 percent.
This rush to the edge comes with concerns. According to Kollective, 66% of “IT teams consider edge computing a risk to the company’s security.” Security across the edge was respondents’ largest worry by far, followed by identifying and deploying edge technologies as well as handling edge-generated data volumes.
Clearly, hyperscale organizations need infrastructure security, and they need it implemented in ways optimized to their specific circumstances. As a global supplier of hyperscale datacenter systems and services, Hyve Solutions now offers a way to do exactly this, with global reach and essentially unlimited scalability, through its Hyve Unified Global Services (HUGS).
Hyve Builds Security Throughout the Stack
Tier-one cloud providers began announcing infrastructure solutions for the edge in 2018. Big names cloned their internal datacenter solutions, offered a limited range of configuration options, and offered to do the systems management. Predictably, these first-generation solutions addressed some needs but left room for improvement.
Obviously, the datacenter is not the edge, and solutions built for one environment do not port cleanly to the other. Just consider that cloud-scale datacenters have the advantage of existing behind robust firewalls and multiple layers of physical security, all within the provider’s IT department’s protective shell. Those factors vanish when you pick up a rack from the datacenter and deliver to an edge facility.
For example, some providers evolved their racks to strip away steel side panels, often replacing them with lightweight acrylic. This helps reduce weight, which in turn makes racks easier to transport and allows more racks within a typical facility. But some providers have proprietary hardware designs and don’t want their IP available for widespread viewing. More importantly, steel is inherently more secure against physical intrusion. Customers won’t want to cut corners on security and put end-user data at risk because of advantages that don’t apply to them.
Hyve took a fresh look at edge security from the infrastructure foundations outward. Knowing that providers have no control over their clients’ physical security, Hyve baked security into every system and rack with a full range of options. Each rack can have its own security camera(s), and unlike with some solutions, those cameras can be custom-tailored in their placement, connection methods, and data characteristics. Racks can have keypads for restricted access. Intrusions can trigger audible alarms and strobe lights. Above all else, every system has root of trust and trusted execution environment technologies, starting at the motherboard. (Hyve even has optimizations available for BIOS and BMC implementations.) In the event of a trust violation, whether physical or via software, owners can opt to have the system’s encryption keys automatically wiped. This immediately puts data beyond the intruder’s grasp and negates the threat of exploit spread.
As with every other infrastructure element, there is no one-size-fits-all aspect to security. Hyve works with clients to customize each security element as needed, starting with a broad set of security building blocks and then fine-tuning them to fit each circumstance.
HUGS: Making Deployment Effortless
If you follow Hyve, you may already be familiar with HUGS and some of its sub-services, such as Hyve’s Rapid Parts Management (RPM). In a nutshell, HUGS specializes in global-scale customized infrastructure deployment and support.
“People normally think about quantity deployment meaning a thousand racks to one location,” says Conor Malone, Hyve’s VP of engineering. “They’re less likely to think of doing one rack to a thousand locations, but that happens all the time, and it requires stringent protocols across a long, secure delivery chain of custody to satisfy customer and government compliance needs. Being able to deliver at scale across those sorts of deployments — and then deliver localized support for them — that’s a very rare, particular set of skills. But that’s what we provide.”
Conventional datacenter-derived offerings from cloud providers will address some needs, including in security and delivery, but not all. One of the largest such holes that Hyve fills is in configuration. An off-the-shelf solution won’t arrive preconfigured. The configuration burden falls on the buyer’s IT department, which may have the expertise for the job but rarely has the time. Hyve’s HUGS program helps ensure that racks arrive in a turnkey fashion, so all that’s needed for racks is placement and plugging in. From there, the solution is ready to run. With infrastructure from other providers, there’s only so much that remote configuration can cover.
“If everything doesn’t bootstrap properly and the security stuff isn’t set up just right,” says Malone, “the customer’s central IT won’t see that system when it turns on. That means either lots of remote service time or someone gets on a plane.”
Hyve processes and services relieve deployers from such burdens. Even orders scaling into thousands of racks scattered around the world will see each system customized, optimized, secure, and ready to run.