Our vision, from inception, has been to build a data driven data center company. We live in a world of data, and we use this data to make physical systems more efficient and effective. We have anti-lock brakes that make us better drivers, converting thousands of sensor readings per second into micro-adjustments of brake pressure. We have self-regulating thermostats that keep us warm. And we have jet engines that are smart enough to extend their own maintenance cycles But we don’t yet have a fully autonomous data driven data center.
A data driven data center company, in theory, sounds great and countless companies will tell you they already do this today. These companies do not. I’ve listened to countless talks about software defined service fabrics and automation of VM’s and containers. I hear marketing terms like “cloud attached,” “data defined,” and “infrastructure as code” but the reality remains that you don’t have an automated or even remotely intelligent data center until you are actually talking to the hardware and the application and making adjustments to both in real-time. For example, it’s not enough to know rack temperature. You can’t do much with that intelligently. But if you couple rack temperature with the cost of cooling and application SLAs, then you can automatically move workloads—based on temperature, cost and business goals—to a location that delivers optimal performance.
When we started Vapor IO we named the company to capture our vision of the data center evolving past large, centralized facilities to a more diaphanous set of resources. In our vision, customers will use our technologies to stitch together facilities of all sizes, from large public cloud providers to micro-regional facilities in cell towers and office parks. After cloud computing came fog computing but that didn’t reasonably represent what we wanted to solve for. We wanted to solve for the mist, the thin gossamer strands of what fog, and mist and clouds were made from. In effect, Vapor. Vapor of course being the gaseous state of something as it transitions from solid to liquid (i.e., clouds) and this accurately provided the right construct for what we’ve been building now for over a year.
At Vapor IO we have created the first real data driven data center. If you need to move a VM or container because of a hardware failure at the server level, we make it so. If you need to migrate a million VMs or containers because of a facility failure, we make it so. If you need to move a container because you have a better place to run it ie closer to the end user/device or otherwise optimized, we inform the data center of that fact and migrate the workload on your behalf. We built the brain that connects the circulatory system (the facility) to the nervous system (the racks and servers).
The world of the data center has been bifurcated into arguably equal but disconnected parts that largely do not communicate. Facility engineers don’t want to know anything about IT and IT engineers don’t want to know anything about the facility. Both take it on faith that their counterpart is efficient and operating well and both are often wrong. We set out to drive telemetry into an interface that both the facility operator and IT operator could understand and interact with. We looked back in time 50 years to the binary wire protocols that run our critical systems (BACNet, ModBus, CAN Bus etc) and turned that data into 21st century REST. We then looked back 20 years to the rack infrastructure layer (PMBus, SMBus, IPMI/IPMB) and turned that data into twenty-first century REST. Then, after providing the world with a common way to describe, catalog and command your data center we made it free for the world to enhance and extend with our OpenDCRE project. Further, we built technology on top of this free software to automate infrastructure (virtual or physical) based on conditions and policies that you create and govern.
The need for data driven data centers will only increase. IoT and Industry 4.0 will undisputedly drive up the amount of machine-to-machine data created and it will drive data centers further towards the edge. It will require us to analyze data and use it to optimize our increasing need for environments that take into account latency, network, cloud, storage, and cost. These changes will usher in a new paradigm of computing where we face new infrastructure challenges. Challenges that Vapor IO is here to help solve.
Lord Kelvin once said,
“To measure is to know, if you can’t measure it you can’t improve it”.
We’re inviting you to build a better data center with us. A data center that can be disaggregated, autonomous, self-healing and policy driven. A data center that perfectly bridges the physical gap between the building and the workload but also the time gap between where we were and where we’re going. This is a different type of cloud. Isn’t it time for a different type of cloud company? Welcome to the world of Vapor IO.