Launched in 2014, Bastille is the leader in enterprise threat detection through software-defined radio. Bastille provides full visibility into the known and unknown mobile, wireless and Internet of Things devices inside an enterprise’s corporate airspace–together known as the Internet of Radios. Through its patented software-defined radio and machine learning technology, Bastille senses, identifies and localizes threats, providing security teams the ability to accurately quantify risk and mitigate airborne threats that could pose a danger to network infrastructure.
Bastille has been a cloud native organization since its inception and its engineers were well versed in the various DevOps practices and services. However, they also realised that infrastructure, while a necessary expense, was orthogonal to delivering customer value. DevOps, itself, isn't the mission of Bastille. This made it the perfect fit for bringing in our external help.
We, like any engineering org, needed to invest in continuous builds, metrics and monitoring. But such infrastructure is a distraction. Our focus should be on advancing our customers and our product.
SuperOrbital worked with Bastille to identify the most important fundamental building blocks of infrastructure that would unlock their velocity: CI/CD and Monitoring & Alerting.
Bastille needed a CI/CD system that they could rely on, that itself was highly secure, and that fit well in an IoT organization. The system SuperOrbital helped implement allows Bastille to run their tests on their own servers, giving them much faster test run time, ensuring their proprietary code and sensitive information stayed behind their firewall, and allowing them to integrate with sensors running locally in their office.
To help move the team toward a test-first mentality, we brought visibility to the build process through a centralized mounted physical build monitor.
Finally, to help with agility and to ensure their deploy technology was continually exercised, we implemented a continual deployment process to a staging environment.
Even as the options for DevOps-related tools improve, it still feels like assembling IKEA furniture. So SuperOrbital made a lot of sense; Tammer could sort out the mix of open source software and hosted tools that were the best fit for our needs and assemble our build, metrics and monitoring systems.
As an IoT company, Bastille faces many challenges around monitoring. The volume of data available to be tracked is immense, and the type of information they have to track to fully understand the health of their systems is highly unique. They needed a strong monitoring and alerting foundation with bespoke tools to collect their distinctive datasets, and they needed to put this information squarely in front of the entire team.
First, we sat down with the team and worked through their needs, helping them identify the dozen high-level metrics that act as business KPIs and to determine where they needed more engineering operational visibility.
Using this as a basis, we helped them understand the landscape of tools, and converge on using SignalFX as the base for their monitoring and alerting system.
We then custom built an agent-based system to collect metrics from their unique data sources. We were careful to ensure this system matched their other internal tools in language, conventions, and deployment practices, making it as easy as possible for the existing team to add new metrics later.
Finally, we organized this data into a useful form, applied analytics, added appropriate alerting thresholds, and displayed it on an information radiator to increase awareness for the entire team.
The knowledge transfer and documentation for my team was easy.
SuperOrbital was the right choice.
As we closed out the engagement, we double-checked our documentation, gave a demo of the system to the team, and walked them through the codebase. It’s important to us to ensure strong continuity of ownership after we disengage.
I’m here to help make sure your next project gets off to the right start.
I embed myself within your team through exploration and implementation. I help identify architectural pitfalls and guide your engineers around them. As an expert in operations and development, I don’t just talk and walk – I grab a shovel and pitch in.
With over a decade of experience running engineering teams of three to three hundred, I also help discover ways yours can be more efficient, effective, and nimble. I've led smaller traditional teams such as RabbitMQ and Redis, and I've also scaled the Pivotal eXtreme Programming process to 50 teams across seven offices in three continents.