Almost every aspect of the technology IT stack is experiencing some form of change. Mobile, IoT and the rise of DevOps have disrupted traditional enterprise architecture and app development methods. Cloud computing, software-defined networks (SDNs), and network function virtualization (NFV) have upended data centers and networking infrastructure. Big data and advances in fields such as machine learning and neural networks are changing how we understand and interact with data.
In every corner, we find opportunity, laced with new and often unknown challenges. One of these obstacles surfaced in a conversation with Evan Kaplan, the Chief Executive Officer of InfluxData, and Tim Hall the company's vice president of Products. The rapid rise of IoT means companies are sensor-enabling vast swaths of equipment. These newly connected devices are producing enormous volumes of time series data, which are data measured over time. Examples of time series data are the measurements you'd get from temperature sensor and accelerometers. While there are many questions in IoT, companies must devise IT strategies that define how to collect, store, visualize and process time-series data. We could use the various Hadoop outgrowths to do this, but is that the best method for this type of data?
Kaplan and InfluxData team believe companies should consider using a method for time series data. The company has developed a four-product platform that it calls the "TICK" stack, named for the four software components that collect, store, visualize, and process time-series data. In October 2017, DB-Engines, which ranks the popularity of data-management technologies, listed InfluxData as a leader in the time series data management category. The same company also noted that the time series data category is the fastest growing in popularity over the past 12 months. However, this portion of the market is still a tiny percentage of the overall database management market.
In truth, I hadn't considered the differences in time series data analysis until I met with the team in June. But apparently, the company uncovered a new market opportunity. InfluxData has more than 70,000 unique open source deployments and more than 300 paying customers. Its customers include Autodesk, Cisco, eBay, AXA, Solar City, Telefonica and others.
However, a company needs more than a good product to be successful. The product has to be available where the developers are, and Influx needs to build partnerships. To that end, the company made two announcements this month to support market expansion and ease deployments. The first was that developers could get started with InfluxCloud directly from the IBM Cloud console, easing the implementation of monitoring on any application or service created within the console. The second announcement stated that customers using Red Hat OpenShift Container Platform for container orchestration and management could more quickly deploy the InfluxData's time series data platform in that environment. According to InfluxData, both IBM Cloud platform and Red Hat OpenShift customers can use InfluxData to build:
● Monitoring, alerting and notification applications supporting their DevOps initiatives
● IoT applications supporting millions of events per second, providing new business value around predictive maintenance and real-time alerting and control
● Real-time analytics applications that are focus on streaming data and anomaly detection.
In our meeting, Kaplan was quick to note that use cases for analyzing time series data ranged from tracking trading activity to analyzing a company's daily sales performance or managing the deluge of data thrown off by a thermostat. The latest announcements with IBM and RedHat signify that InfluxData is working hard be present in the right markets. The question now is will customers embrace another data management method. With a wide range of use cases that span numerous industries and the rapid IoT enablement of the world, there's likely plenty of opportunity for time series data management.
