When Amazon launched the world’s first true public cloud service Amazon Web Services (AWS) in March 2006, very few people recognized how much it would change enterprise IT. AWS had come out of nowhere and, initially, none of the major infrastructure vendors took it seriously. After all, Amazon was an online retailer. What did they know about enterprise infrastructure?
As it turns out, Amazon knew a lot. By the time AWS’ rivals recognized that the cloud wasn’t just a sandbox for developers, it was becoming a threat to their enterprise-focused businesses. And at that point, Amazon had four years of experience building, running and delivering hyperscale cloud services. Microsoft wouldn’t launch Azure until February 2010, and in November 2011, Google officially released App Engine, which would later become Google Cloud.
As an entrepreneur building a cloud company in 2008 with my co-founder, John Considine, I had front-row seats to this cloud transformation and its potential. At the first AWS re:Invent conference in 2012, although S3 and EC2 were already six years old, it still felt like an insider event, filled with the energy and wild-west confusion that accompany a new market. At that time, the rising role of developers coupled with early test and dev use cases fueled exploration and adoption of the cloud. At the same time, enterprises observed the progress waiting to see if and how the cloud would transform infrastructure as they knew it.
Amazon, Google and Microsoft weren’t the only cloud games in town by then. Colocation providers, such as Terremark, Rackspace and Savvis, tried to become cloud companies, but in the end, couldn’t achieve the hyperscale they needed to succeed. Carriers like Verizon and CenturyLink also tried to provide cloud services, but they were technologically too far behind the leaders to catch up to the global scale that AWS, Azure and Google had achieved.
Other colo providers, such as Equinix, took a different path. They focused on the value they could provide by virtue of having highly connected co-location facilities close to customers. It was a smart move, because it turns out that the hyperscale cloud has an Achilles’ Heel — latency.
The three big public cloud providers built their enormous facilities in sparsely populated areas where real estate was cheap, and often hundreds or even thousands of miles from the metro locations where most of their customers reside. These distances introduce unavoidable latency because, as everyone knows, you just can’t change the speed of light.
ClearSky: Early at the edge
When Laz and I founded ClearSky in 2014, we knew the cloud had enormous potential to change enterprise data storage, but to realize that potential, we had to overcome the latency problem. It was extremely early to be thinking about the edge.
After all, most enterprises were just starting to consider putting critical apps in the cloud. Edge computing was not yet on the horizon, but ClearSky was already there, building a service architecture that included the edge from the very beginning.
Having lived through the early development of the cloud, I’m happy to see the edge is evolving much faster than the cloud did. While the cloud was conceived with a vision for new use cases and infrastructure opportunities, the acceleration of edge computing is fueled by persistent demand for high-performance access to data generated from sources that go far beyond the traditional data center or cloud.
The hyperscale cloud players, in their determination for everything to move to the cloud, dismissed the edge and only recently started rolling out their edge services, like AWS Outposts, Google Stadia, Azure SQL Database Edge and Azure Stack.
Colo providers like Equinix, who tightly interconnected its more than 200 facilities around the world both to one another and to public clouds, are now perfectly positioned to provide edge services. And they’re not the only ones.
Carriers are repositioning and refining their metro facilities to provide edge service, just as mobile carriers are finding new edge roles for cell towers. Both are well positioned to deliver edge services across different geographies to support the exploding number of edge locations.
The strong demand for edge computing is even spurring the creation of new metro micro-cloud providers like MetroEdge, who is building facilities in underserved communities in large cities, and micro-modular data center providers like Vapor IO and EdgeConneX. At the same time, interconnection and telco providers including Packet, SBA and StackPath, are growing their businesses based on edge-computing needs.
A union of the cloud and the edge
Today’s reality is that the edge represents a paradigm shift in the way we design, build and deliver IT services, and a new edge ecosystem is emerging. The leaders noted above and ClearSky are at the forefront of developing it.
Because the edge is growing so rapidly and the opportunity is so large, there’s even more danger of incumbent storage and IT infrastructure vendors engaging in “edge washing” than there was for “cloud washing” earlier this decade. There’s a lot more to the edge than plunking down some hardware in a colo, as Thomas Bittman, VP and distinguished analyst at Gartner, wrote in his 2017 blog post, “The Edge will Eat the Cloud”:
It’s one thing to have local processing next to a number of fixed devices, monitoring pipeline pressure, etc. – it’s another when those things are in flight, moving constantly, and perhaps even appearing – digitally – at a location.
… All of this requires processing power and storage, lots of it. And data analytics tools. And tools to push software and data to the edge. And tools to federate across the edge, and with the centralized cloud. And machine learning on the edge itself.
The edge will need some serious muscle.
Simply put, the edge is developing so quickly because the industry clearly recognizes that the cloud needs the edge. There’s no other way to provide the performance and quick response for emerging use cases that require analytics, content distribution and fast processing of data. AI-powered apps that are generating data across many disparate sources, in real time, are driving the need for edge-based solutions.
But the inverse is also true: the edge needs the cloud. IoT, enterprise storage, connected vehicles, smart cities — these use cases will produce enormous amounts of valuable data that cannot be stored forever at the edge. The cost and footprint of traditional storage models make no sense in this new paradigm.
The cloud is massively redundant with essentially unlimited scalability and capacity. There’s no better place to store, analyze and work with such gargantuan volumes of data – so long as latency doesn’t pose an issue.
In November, we announced a partnership with Equinix. Not only does this partnership let us rapidly expand into new markets, but also gives our customers an opportunity to take advantage of the Equinix platform and its direct connections to multiple clouds.
We’re also actively collaborating with other edge players to create a mutually beneficial edge ecosystem. ClearSky’s data fabric provides direct access to customers who aren’t set up within their data centers, extending their reach all the way to the customer site. For ClearSky, we can leverage their facilities to expand and augment our service offerings.
The edge unlocks the full potential of the cloud, enabling enterprise IT to radically simplify, augment and optimize core infrastructure that was previously confined to on-prem facilities. We’ve barely scratched the surface of the possibilities.
Interested in learning how ClearSky combines the edge with the cloud to provide a complete enterprise storage service? Sign up for a demo today.