Update how internal network is configured

[longshuicy]

Need to clean up later based on the discussion:

• Centralized place in global.py to set all urls
• Allow a file path that holds all the configurations; load file to populate the urls
• Leverage helm chart config map to dynamically mount that file to destination on datawolf pod
  • Difficulty is how can datawolf tool pods acquire that file; there isn't any specific mounting in the helm chart for tools. Instead the k8s executors are in charge of spin off the tool pods
  • Alternatively, can use the k8s executors's dockerfile to write config file in place (?)