Why Log Management is a Key Underpinning of the Internet of Things
There’s a lot of buzz surrounding IoT devices, but the success of this emerging sector largely depends upon its backend development.
While the Internet of Things (IoT) has been in development for a number of years, it’s fast becoming the technology everyone is talking about, as evidenced by its tremendous presence at this year’s Consumer Electronics Show.
For those unfamiliar with the IoT, it is the ability to connect, communicate with, and remotely manage networked and automated devices via the Internet. This includes technologies both at home as well as those at manufacturing facilities, company offices, and more.
Industry leaders predict that the number of “connected” devices will surpass 15 billion by 2015 and 50 billion by 2050. In terms of dollars and cents, venture capitalists invested more than $1 billion dollars in the IoT space in 2013, and experts predict that the IoT will boost the global economy by $1.9 trillion come 2020.
While most of the buzz surrounding this sector has to do with which gadget and device is IoT-capable, including lights that turn on in the home as one pulls up the driveway, thermostats that automatically turn down when no one is around, and more, there are matters that developers and operators need to consider with this technology. Specifically, how they will be able to diagnose and solve the problems that will inevitably pop up with these devices. That is, operators must find efficient ways to effectively troubleshoot issues remotely, keep downtime to a minimum, and manage the billions of log events these devices will generate.
At present, supporting frameworks for the IoT are still in the early stages of development. Fortunately, there are groups like the Open Interconnect Consortium (OIC), which was founded this past July by some of the world’s leading technology companies. These companies are jointly taking on the responsibility of defining the connectivity requirements and ensuring the interoperability of the billions of devices that will make up the IoT. Standards could cover IP protection and branding for certified devices and service-level operability, and allow for application developers to work seamlessly across Android, iOS, Windows, Linux, Tizen, and more.
Cryptosoft is one of the industry players to have formally joined the OIC. In a news release, Jon Penney, the company’s CTO, explained: “The broad acceptance by consumers and the industry will require a simple and open framework that assures secure interoperability of data flows between the billions of devices anticipated to make up the Internet of Things.”
While the OIC is focused on issues related to making the IoT come to life, there has been little discussion about standards for addressing the aforementioned operational issues associated with the IoT. Specifically, how DevOps teams can manage IoT logs and address the inevitable issues these devices will experience in an efficient and timely manner.
Syslog is one simple logging protocol that could easily extend into the IoT world. Because it’s very well spec’ed and understood, adding support for it to an IoT device should be straightforward, even for the simplest and least powerful devices. Some IoT devices may also need REST APIs for more sophisticated logging. (Fortunately, Loggly supports both syslog and RESTful log transmission via HTTP :-).)
Logging is moving away from simple text based data to more structured formats, and the line between “data” and “logs” is blurring as log management systems get better at dealing with this new type of data. JSON is a simple but powerful serialization format that can be used today to push structured data (performance metrics, sensor readings, etc) to smart consumers. Your logging platform can be one of those consumers.
Regardless of the outcome of standards discussions, the most important thing to consider is that with all of these advancements being made, IoT DevOps teams will need a highly efficient log management system to help them quickly diagnose and resolve issues that might occur at any given moment, much like those already in place for net-centric operations like Software-as-a-Service, online games, and other digital businesses.
Access to a sophisticated cloud monitoring platform boosts Ops team productivity by preventing / reducing downtime when possible, and ensuring faster resolution of issues when they do occur. This includes monitoring of all critical devices in a particular IoT network, endpoints and services, release monitoring, notification alerts of potential issues, data reporting, and more.
Such efficient monitoring software not only ensures constant IoT / network uptime, but it’s cost effective too in the sense that issues are dealt with proactively, not reactively.