The state of data at the edge is … there’s a ton of data being generated, it needs to be secured and managed, and that requires orchestration and automation.
This, according to the first topic-specific edge report published by Vapor IO, Packet, and Seagate Technology. The free report follows the inaugural “State of the Edge” report published last June.
A group of vendors —Packet, Vapor IO, Rafay Systems, Ericsson UDN, and Arm — came together to work on the earlier report because they believed that vendor-neutral research on edge computing was sorely lacking. That initial report also included an open glossary of terms related to edge computing. “The glossary was one of the most popular pieces, and it was so popular that the Linux Foundation adopted it as a project,” said Vapor IO Chief Marketing Officer Matt Trifiro, who co-chaired the State of the Edge report.
In January, the “Open Glossary of Edge Computing” became part of the Linux Foundation’s new LF Edge organization. And today, the State of the Edge group donated the Edge Computing Landscape Map to the Linux Foundation, to be led by the Open Glossary project.
A couple other significant things came out of the vendor-neutral research project. “It was very well received, and there were two things people asked for,” Trifiro said. “They really appreciated this landmark report once a year, but they also wanted topic-specific reports. And lots of other companies wanted to become members and participate.”
So the group decided to produce more topic-specific reports — the first is today’s “Data at the Edge: Managing and Activating Information in a Distributed World.” And they decide to move from a sponsored organization to a membership organization.
State of the Edge Membership
“We wanted to democratize membership,” Trifiro said. “Companies that want to join pay a $7,500 annual membership fee, and 100 percent goes to funding this open research. All the research is distributed under a creative commons license. There’s no commercial goal here other than to accelerate the industry and this rising-tide philosophy.”
Startups with 30 or fewer employees get a discount. They pay $2,500 per year. Trifiro expects to count 40 member companies by year-end. In addition to the big State of the Edge report, the group also plans to publish about five other topic-specific reports this year.
As the name suggests, the new report focuses on managing and activating information using edge computing. It finds data is proliferating at unprecedented speeds. Data generated by 2025 is expected to exceed 175 zettabytes, a tenfold increase from 2016 levels. And the need to manage this staggering volume of data is going to be a key driver of distributed architecture.
In addition to the high data volume accompanied by insufficient bandwidth, other factors driving demand for edge computing will be latency, cost, data sovereignty, and compliance.
All of this data being generated, used, and transmitted at the edge requires management, and the report says this will depend on two technologies: orchestration and automation.
“This is an immense amount of data,” Trifiro said. “And then you go one level beneath it: this data is being generated by machines and intercepted by machines, so we’re building this next extension of the internet though the edge and edge clouds.”
Data Security at the Edge
The report also says that three things, in particular, are unique to data security at the edge. As devices enter and exit the network, they must be trusted. Additionally, the data itself must be trusted, whether used immediately in analytics and artificial intelligence (AI) workloads, transmitted across nodes, or uploaded to the cloud. Finally, data in motion will be more vulnerable as it traverses networks, so nodes need to communicate securely over disparate networks, the report says.
“There’s the physical security concern,” Trifiro added. “Data centers are potentially anywhere and they can’t necessarily have people at them. So you need to have a pretty sophisticated approach to physical security. A lot of edge data centers are also multi-tenant and one of the challenges is how to you protect company A’s equipment from company B’s equipment.”
Vapor IO rotates the racks in its data centers to solve this problem, he said, and other cloud providers take different approaches.
In addition to physical security concerns, there’s also the threat of software vulnerabilities at the edge. This highly disaggregated environment means a greater surface area for potential attacks. Encryption and authentication technologies become more important at the edge to protect data and establish a trusted network of devices and data. AI and machine learning also provide another way to secure the edge and its data.
The bottom line when it comes to securing data at the edge: “People in the security business are going to have a field day,” Trifiro said.