We all hate latency. Here’s what some IT departments are doing about it.
Sometimes latency is a matter of life or death.
That’s what first responders in Arizona’s Maricopa County realized when their 911 system couldn’t supply crucial data quickly enough. When an emergency call came through, operators updated a cloud-based system with details on where the emergency was, who was involved, and what was wrong.
But police and firefighters, trying to access that data through an app on their phones, had to wait. The county’s legacy infrastructure couldn’t support the influx of data, and failed to push it out to endpoint users quickly.
This is one of IT’s largest challenges: finding ways to leverage existing infrastructure with new cloud technologies, using those investments to meet demand and maintain efficiency.
And it’s not just in the public sector. Southern New Hampshire University (SNHU) struggled to support the cloud-based applications students wanted to use. In addition, it still employed aging in-house apps to report grades to students, who wanted easier access that existing systems couldn’t handle. As SNHU pursued a hybrid model of on-campus and online education, the need for better infrastructure and faster access would only grow.
In both these cases, close integration of existing infrastructure with the cloud was achievable. IT is defeating latency with converged infrastructure, an option that would deliver relevant 911 data as it’s available and give students the technology resources they need to get the most out of their education. Here’s how.
A latency emergency for 911 responders
Maricopa County’s struggles with a sluggish 911 system stemmed in part from the desire to leverage its existing system — an understandable attempt to extend its longevity. But the speed of change in data demands meant the hardware couldn’t keep up.
Ultimately, county officials invested in a back-end system capable of supporting a hybrid cloud environment. This IT improvement meant that up-to-date information was delivered in seconds, and latency disappeared; the hardware better supported the cloud-based application, and the entire system could better deliver critical information to the first responders in the field.
Fixing slowdowns with a hybrid cloud
This example teaches us that a fully integrated system is necessary to truly run an efficient hybrid cloud solution. But most systems aren’t integrated because each component is purchased individually; when adding complete systems to a data center, every piece of hardware (storage, servers, and networking) and software is built by different vendors, which will inevitably lead to inefficiencies. That often creates latency.
Think of it this way: When you buy a car, you don’t buy each individual part. You go to the dealer for the complete vehicle. Why should your IT infrastructure be any different?
A converged infrastructure assembles those assets – server, storage component, network components, and often software – in one device to allow for better integration of cloud-based applications. A close integration should lead to better efficiency and data deployment in real time. Because converged infrastructure devices are built by one vendor, they’re unlikely to suffer from some of the inefficiencies of disparate components.
Keeping latency away for good
SNHU followed this playbook to create a digital backpack in the cloud that served students both on and off campus, allowing them to access work, collaborate with their peers, and access lectures from any classroom, dorm room, coffee shop, or library. Students could use the apps they wanted. Latency had been eliminated.
Besides reducing latency and increasing efficiency, there are a number of other reasons why a converged infrastructure might make sense. For one, it can reduce maintenance costs for your IT team – after all, having one vendor to go through is easier than three. That’s a layer of complexity that’s eliminated just by using one converged infrastructure unit.
This one device also offers better visibility into the back end of your system, so you’ll have a better understanding of just how efficient the environment is. An example of this would be virtual machines – with a converged unit, it’s likely easier for IT to understand how many virtual machines are being used on the system. Often, we see organizations paying for more virtual machines than they need, in part because they don’t have good visibility into their IT infrastructure; a converged infrastructure unit eliminates that oversight.
In IT’s eternal game of catch up, where the next improvement is only months away, can we win with converged infrastructure? Maybe. One thing is for certain: It will improve your support of cloud applications while keeping the overall IT environment more efficient and agile.
Jim Stevens manages HPE’s Hybrid IT business with National Solution Providers (NSPs). He has over 20 years’ experience in business development and sales management roles for various IT startups and enterprise organizations, including HP, Opalis (acquired by Microsoft), Virtual Computer (acquired by Citrix), and DynamicOps (acquired by VMware). Jim specializes in virtualization, cloud computing, converged solutions, management and automation/orchestration software.