Edge Computing Security Dos and Don'ts

Edge computing can significantly improve the efficiency of collecting, processing, and analyzing data gathered by arrays of IoT devices. Yet to protect the IoT deployment, as well as the overall enterprise network, it's essential to implement complete and reliable edge computing security.

The security risks associated with edge computing are very different from the security risks associated with a centralized environment, like the cloud, observed Duncan Pauly, CTO at Edge Intelligence, a computing analytics company. "With the cloud, all of your data will likely be in a single or small number of locations," he explained. "By comparison, with edge computing, data is decentralized, which makes it much more challenging to have a complete dataset compromised."

(Image: Shutterstock)

Ensuring edge computing security is much more challenging than providing cloud security due to the fact that edge computing involves distributed data processing. "There are a lot of devices—usually various sensors and controllers—that work differently," said Leigh-Anne Galloway, cybersecurity resilience lead at Positive Technologies, an enterprise security solutions provider. "Each [device] is configured in its own way, meaning there will be different versions with different vulnerabilities, and this causes many problems." Galloway noted that while edge computing is a relatively new technology, old problems remain, including weak log-in credentials, zero-day vulnerabilities, a lack of updates and a less than optimal network architecture.

Many IoT adopters mistakenly believe that edge technology inherits the same security controls, processes, and checkpoints that are found in private data centers and the public cloud. "The reality is that physical edges reside in all sorts of environments and are often remotely managed and monitored," said Haseeb Budhani, CEO and co-founder of Rafay Systems, a programmable edge company that's set to begin operation early next year. "An edge may not be as secure or reliable as the public cloud environment organizations are used to," he warned. "Edge customers must be vigilant with respect to vetting vendors as to their security architectures and practices, using public cloud environments as the security standard," he added.

Best practices

Shane MacDougall, senior security engineer at networking and cybersecurity services company Mosaic451, observed that the best security step any edge computing adopter can take is to equip all edge nodes with the same level of protection as the rest of the network. "It is critical to remember that the security of your network is only as strong as your weakest link," he remarked. "Therefore, ensuring that every host is hardened and patched is critical to maintaining a secured environment."

Dennis Chow, vice president and information security advisory leader at SCIS Security, a Houston-based network security firm, stressed the need for continuous monitoring and visibility. "Prevention is preferred, detection is a must," he stated. "[Activities] may include decrypted network traffic, logs, and other near-real-time monitoring capabilities."

Pauly stated that edge computing users also need to ensure that all data, both in-flight and at-rest, be encrypted, and that all communications within an edge computing environment utilize SSL/TLS security along with multi-factor authentication access. "If done right, [this approach] diminishes the importance for physical security, as the data is encrypted with secure access in-place."

It's also important to keep in mind that edge computing will increasingly be deployed in places other than data centers, where physical security cannot be guaranteed. "Secure data encryption and secure access authentication will be ever-more essential in those environments," Pauly observed.

Worst practices

One of the biggest mistakes edge computing users make is assuming that traditional security controls alone will provide adequate protection for their devices. "For example, using [only] antivirus and firewalls to keep their organizations safe from cyberattacks likely means they are already a victim of cybercrime, but just have not found [the evidence] yet," said Joseph Carson, chief security scientist at Thycotic, a privileged access management technology provider. "The organization’s data is no longer flowing through their Internet connection, nor via their corporate firewalls, so they must need to secure and protect each edge device today as if it is a door to their network," he warned.

Another mistake often made by edge newcomers is simply applying cloud-based security models to edge computing deployments. "AWS and Google have 20,000 engineers, each securing 56 physical locations (in AWS's case)," observed Ian Eyberg, CEO of NanoVMs, a virtual machine security technology supplier. "If you're a retailer, and you have 2,000 locations that you want to run edge compute at—like a Chick-fil-A is doing now—you have 2,000 locations that need the same security, and chances are you don't have 20,000 engineers at your disposal like the big public clouds do."

Final thought

"Edge compute is going to usher in a completely new reality of what we consider cyber security and cloud infrastructure, not to mention all the benefits edge brings," Eyberg concluded. "Unfortunately, I think we are going to have to learn some hard lessons first."