The problem they run into is that, while they're clearly the experts at fixing one-off issues, it becomes really challenging to deal with all the distributed systems concerns.
Examples:
- What do you do when part of your network is unavailable, limiting your visibility to see what's happening, much less fix it?
- How would you prevent the automation from running amuck and creating a bigger problem?
- How do you scale the solution as your fleet grows?
- How do you deal with data in many silos for resources, metrics, etc.
- all slightly out-of-date and inconsistent?
These are the problems we have set out to solve for our customers. Our platform is:
- real-time and fault-tolerant throughout
- operates at the edge, so you scale with the fleet
- operates locally, so you don't need a single central point of failure to be up
- able to provide circuit breakers to prevent things from going out of control
The customer is responsible for figuring out how to make the change on the individual box, but we do all of the orchestration around it.