3 years ago

Category: Data, Kubernetes

Coming along the road to battle between Dev and Ops, earlier we used to face tussles in between Dev and QA, Many thanks to DDD which helped to mitigate that gap, These days the heat is more on the conversations in between Dev and Ops related to automation during CI/CD or provisioning the infrastructure and taking charge of configurations and deployment processes. Is it the developers who must be responsible and keep the configuration management at their end or is it the Ops who must maintain and take care of it”

Lot of developer teams are quite confident on managing the configurations and bundling the configurations as part of their jars using Spring cloud config server(actually a good way to manage configs) and then activating the profile in bootstrap or passing profile in VM arguments while deploying the application.

To my experience the choice is between better management and better process, As with the developers handling all the configurations it makes more sense in terms of less dependency, less resourcing, less processes, less hassles with config.

If there is not a lot of gap in between the teams and if communication channels and processes have been well established then it does not really makes a lot of difference as to who is handling the configurations( This is obviously in theory).