Developer Operations

accelerating the development lifecycle to deliver hardened and quality solutions

Project Details

  • Focus : Cloud, Automation
  • Research : Engineering Practices
  • Use Case : Accelerating software delivery
  • Use Case : Updating legacy systems

Developer Operations

The traditional Waterfall development of systems sets up a serial process that makes changing requirements difficult and expensive after a contract is awarded. Agile development processes, such as Scrum, have allowed iterative development with more flexible requirements, and is a process Reallaer has implemented and ran with for many years now. During this time, we also made the transition to hosting many of our services within the Amazon Web Services infrastructures. When clients were muscling for time on high performance computing clusters, we shortened the process by migrating their processes to scalable cloud infrastructure. To improve developer response time and increase test coverage, we jumped on containerization technologies. Reallaer continues to introduce these concepts to our clients, bringing new tools that make an impact on schedules and maximizing the quality and flexibility of our solutions.

In August of 2019, the Department of Defense released their Enterprise DevSecOps Reference Designs. Realizing the need to update their acquisition and development practices, the DoD has begun migrating their software practices to developing at the speed of operations. Many of these designs are tailored to software development and security practices baked in from the start of the development. There is a pressing need for people not only experienced in developer operations concepts, but also for those who have the experience in government domains.

Simulation Operations

Our current research dives into emulating the developer operations practices within simulation domains. These domains often meet chokepoints in the time it takes to execute change requests, to test application/environment interactions, or sometimes simply running the simulation as a job on a local cluster. The acquisition process is overall slowed as systems fall into a trial and error process. The use of simulations is useful for a number of reasons and cannot live as legacy systems going forward if they want to respond to domain changes that continue to accelerate.

We have named this area of research, Simulation Operations. There is no easy and simple solution to adapt these evolving environments. Large entities are challenging to exhaustively model, and individual environments can be difficult demand many resources to stand up. Adopting methods that allow for composable environments will improve our ability to respond to either domain or requirements changes. Working with cloud vendors to secure infrastructure that is compatible with our client’s security needs is a constant challenge and one we are navigating as vendors such as Amazon rolls out their federal cloud solutions. Our goal is to provide the benefit of developer operations within a controlled modular/composable simulation environment that is useful across many phases of a system’s lifecycle.