Make dump output more easily diffable

[janv8000]

The YAML output produced by a dump does not always retain the same order (between runs or between different instances).

Output 1

host: <redacted>
https: true
headers:
  - 'X-APIKEY:...'
apis:
  - name: api1
    plugins:
      - name: second_plugin
        attributes:
          enabled: true
      - name: acl
        attributes:
          enabled: true
  - name: api2
    plugins:
      - name: acl
        attributes:
          enabled: true

Output 2

host: <redacted>
https: true
headers:
  - 'X-APIKEY:...'
apis:
  - name: api2
    plugins:
      - name: acl
        attributes:
          enabled: true
  - name: api1
    plugins:
      - name: acl
        attributes:
          enabled: true    
      - name: second_plugin
        attributes:
          enabled: true

The two outputs are functionally equivalent but comparing them is a bit more work.

Ideally api's would be ordered by name/id, then plugins by name/id, then ...

[CyExy]

This would fit in the try-run category of features, or in other words diffing the server state and your config file and getting an output that would show what would change if you would apply your config.

To understand more your reasoning and use-case; why is it useful to you to compare the outputs?

[janv8000]

We’ve got 3 environments with separate Kong’s running (dev, QA, production). Setup and configuration is mostly version controlled and automated. Being able to compare two Kong instances/environments however would enable spotting configuration drift … Besides that in DEV we play around with the Kong configuration using https://github.com/PGBI/kong-dashboard, comparing the before and after would help in automating those changes...

…

Op 6 feb. 2018, om 21:59 heeft Sten Hiedel ***@***.***> het volgende geschreven: This would fit in the try-run category of features, or in other words diffing the server state and your config file and getting an output that would show what would change if you would apply your config. To understand more your reasoning and use-case; why is it useful to you to compare the outputs? — You are receiving this because you authored the thread. Reply to this email directly, view it on GitHub, or mute the thread.