v0.4.0

Added

• #17: Ability to host the /status endpoint on either (or both) the main (default :8050) or the diagnostics (default :8051) HTTP ports. By default, it's enabled for the diagnostics port (default :8051) only.
  • Enabling for the main HTTP port:

    http:
      # port: 8050
      status:
        enabled: true

  • Enabling for the diagnostics HTTP port:

    diag:
      # port: 8051
      status:
        enabled: true 

Fixed

• #18: When a non-existing feature flag key is posted to the /api/{sdkId}/eval endpoint, the response is now a 400 Bad request instead of the 500 Internal server error.

Breaking Changes

• By default, the /status endpoint is not enabled anymore on the main HTTP port (default :8050). Instead, if not configured otherwise, it's only available on the diagnostics port (default :8051).
• The /metrics endpoint configuration (along with the /status endpoint's configuration) is moved under the diag configuration node.
  Before:

  metrics: 
    port: 8051
    enabled: true

  Now:

  diag:
    port: 8051 #<-- port of the diagnostics webserver (that will host the private '/metrics' and '/status' endpoints)
    enabled: true #<-- controls whether the diagnostics webserver should run or not
    status:
      enabled: true #<-- controls whether the '/status' endpoint should be enabled on the diagnostics HTTP (8051) port
    metrics:
      enabled: true #<--controls whether the '/metrics' endpoint for Prometheus should be enabled on the diagnostics HTTP (8051) port 

  The names of the metrics related environment variables were also changed:
  • CONFIGCAT_METRICS_ENABLED => CONFIGCAT_DIAG_METRICS_ENABLED
  • CONFIGCAT_METRICS_PORT => CONFIGCAT_DIAG_PORT