Organizations rolling out the IoT usually aren’t prepared for the additional complexity. With the IoT, data volumes grow exponentially, infrastructure management gets more complicated and the security vulnerabilities increase disproportionately. Nevertheless, IT departments are expected to handle all these changes competently without proportional increases in budget or other resources.
With Red Hat’s expert assistance, IT and software organizations can manage IoT adoption with greater ease, so they can spend more time delivering value and less time recovering from common pitfalls that could have been avoided.
“Enterprises sometimes manage enterprise systems one way and IoT another,” said Ishu Verma, IoT technology evangelist at Red Hat. “If you’re approaching those things differently, you’re not managing the data, security or your resources as well as you could.”
Get IoT data under control
In today’s data-intensive business environments, some organizations want to save all data because storage is cheap and they don’t want to discard something that may be valuable.
IoT devices generate a lot of redundant data, however. In most cases, status changes and other outliers are more valuable than 10,000 pieces of static, repetitive data because the behavior outside the normal signals the need for action. Using a smart algorithm like sliding window, most of this sensor data can be summarized into a more manageable size.
“As more things get instrumented and you gather more data, you may find that the volume of data is growing faster than you can manage it, so you want to make intelligent decisions about data velocity at the edge,” said Verma.
IoT gateways monitor streaming IoT data and then make intelligent decisions about what data the enterprise should receive. That way, an organization can monitor all the data while incurring the costs of storing only the most meaningful data. By reducing the amount of data to be analyzed, decisions can be made in near real time, a key requirement for critical equipment.
IoT gateways can be designed and implemented as hardware components or virtualized. Linux containers provide an elegant solution to manage IoT data because they can be provisioned automatically to scale as the volume of data requires. Containers also provide security capabilities to segregate critical and non-critical data and devices.
Keep IoT data secure
Hackers increasingly target industrial equipment because the security of the devices has not been addressed adequately. In many cases, the root cause of the vulnerabilities is the failure to patch or otherwise update the operat- ing system. By the time the vulnerability has been identified, important systems have been compromised via a lateral or denial of service attack.
“Security is a complex problem. You need a multi-layered approach that includes physically securing the sys- tems, pre-boot authentication and an operating system with security capabilities like SELinux to limit access to system resources and data, and data security at rest and transit,” said Verma.
“Each of the layers has to be implemented correctly.”
The low-end devices and sensors present the biggest security risks as demonstrated by recent exploits including the Mirai botnet. These network-capable devices lack the security implementation of a typical enterprise system. From design point of view, low sensor costs may not justify the inclusion and maintenance of an operating system. An IoT gateway can provide a firewall to protect the low-end devices and sensors so they can’t be accessed directly from the internet.
“It’s important to secure the end devices, but you also have to ensure that the gateway is secure. People miss that,” said Verma. “An API management system should be part of your security stack.”
Rationalize IT and IoT
Resource constraints tend to worsen as the technology stack becomes more complex, but getting proportionate funding and resources is out of the question. By aligning enterprise and IoT efforts, businesses can better leverage the resources they have and be more productive using modern DevOps techniques.
“IoT adoption hits a wall when you lack the skills you need,” said Verma. “We recommend using an open-source solution built from cloud to the edge using the same tools and processes for both IT and IoT.”
Learn more at www.redhat.com.