The Reliably CLI uses rules to surface potential reliability weaknesses in your resource declaration files or in running resources.
How Rules work
Reliably relies on sets of rules to identify patterns in your resources that might result in unreliable behaviour for your system. When scanning your resources, the CLI fetches the requested sets of rules and runs them against your resource. This ensures all the analysis is done in your own premises and your code never leaves the safety of your environment.
Where the rules come from
Reliably Rules are organized in sets of rules, sorted by platform and kind of resources. For example, we have a set of Kubernetes-related rules, which comprises subsets of rules for Pods, Deployments, etc.
The rules have been written by the Reliably team and reflect best practices, Kubernetes official recommendations, and our own experience configuring and running distributed systems.
We're investigationg letting users submit rules that they think should trigger suggestions. If you want to know when this is put in place, follow us on Twitter.
Are this rules mandatory?
Sometimes, you might be in a situation when you know you're going against the general consensus for some good reasons. Feel free to overlook them if you know what you're doing!
We're working on a way to allow users to prevent some rules to be triggered, on a file or line-basis, so that suggestions you know you will overlook won't appear. As all workarounds, you will want to use it with caution.