Sea Street Technologies wants to give you an operating system for everything. And it does mean everything: infrastructure management that doesn't care what your infrastructure is or where it
The 451 Take
Most cloud and IT orchestration models are based in scripting techniques, which (for example) are triggered by events or run on timetables – logging and tracking are separate processes. Sea Street puts the required management actions and the impact they have
This extends to mobile devices, factory equipment that is computer-controlled, transmitters, sensors, even things like facility controls and computer-equipped generators. That said, there's a lot of infrastructure management software out there, and some of it can make similar claims to automation and orchestration. What makes
The platform is built on a top-down, loosely coupled hierarchal object model, with each object containing defined controls, objectives and instructions. The top of the chain defines the overall objective (delivering an IT service) with parameters, metrics
Down the chain from this service model are smaller and smaller objects controlling and receiving input from various parts of the IT stack that comprise what's necessary for service delivery. A simple example would be delivering a mobile app for a public cloud to a set of endpoints. The service delivery is defined as 'the devices are receiving appropriate traffic from the app.'
Objects are created for each of the infrastructure components (public clouds, mobile endpoints, the application and networking). As StratOS monitors these components, it detects problems and takes action as defined in each segment, ultimately reporting back to the top service-delivery object. Resiliency is created to the required level by replicating objects across redundant infrastructures; they bring all their instructions and information with them. Sea Street calls this an 'atomic architecture,' with smaller state machines orbiting the nucleus of the governing service-delivery object.
Some of the signal advantages to this approach are that it is truly platform- and hardware-agnostic. It can be extended across modern cloud platforms and 30-year-old legacy telecom gear at the same time with positive results. It can be set up to be as simple or as complex in terms of operations, workflow and execution as needed. It can address every facet of the IT organization, from locating and ordering replacement of failed hardware to software and OS maintenance to performance optimization and compliance – anything that can be described in a workflow can be placed in an object and added to the state machines in play.
Changes to objects can be made directly in the command line, and take place in real time. To a limited extent, it can be taught to learn and improve on its own; the property of stateful operations and a feedback loop and response effectively makes this a software robot capable of acting autonomously.
ContextOverall, Sea Street's approach is a remarkable departure from how most IT operators think about systems management, especially at the large service providers the company is targeting. It is precisely these kinds of organization that stand to benefit from this approach. The downside is that the learning curve is steep for those traditional operators, and the idea of levering management tools all the way out of the data and into this abstract, yet functional,
Once in place, however,
This can even be an explicitly defined objective; and given the proper instructions,
Notably, there is a minor trend in using state machines to manage container fleets, specifically with Apache Mesos, and a stateful approach to data management that employs the same kinds of stateful feedback loops that Sea Street
It is also the case that large cloud providers like Google, Facebook and AWS employ machine learning to improve their operations. However, no other vendor has made the configurable state machine the primary mechanism of management and
Carl Brooks is an Analyst for 451 Research's Service Providers Channel, covering cloud computing and the next generation of IT infrastructure. Previously, he spent several years researching and reporting on the emerging cloud market for TechTarget.
Jeremy Korn is a Research Associate at 451 Research. He graduated from Brown University with a BA in Biology and East Asian Studies and received
Aaron Sherrill is a Senior Analyst for 451 Research covering emerging trends, innovation