kie-tools#2695: [SonataFlow Management Console Image] Use mod_proxy to proxy request to the data-index

[fantonangeli]

Closes #2695

Description:
Currently, the SonataFlow Management Console attempts to trigger workflows using internal Endpoint URLs returned by the Data Index (DI). These URLs are not accessible from outside the cluster, leading to errors when users attempt to initiate workflows through the console.

Please configure Apache Httpd in the Image using mod_proxy to expose the Data-Index outside of the cluster, doing the communication on the server-side.

Preview:
kie-tools-issues#2695.webm

How to test:

• Run the Docker Image Locally from packages/sonataflow-management-console-image

  1. This command will start the container and configure it to access your local data-index service from the container.
     Replace <HOST_IP_ADDRESS> with the IP address of your host machine.

     docker run --rm -it -p 8080:8080 \
       -e SONATAFLOW_MANAGEMENT_CONSOLE_KOGITO_ENV_MODE='DEV' \
       -e SONATAFLOW_MANAGEMENT_CONSOLE_DATA_INDEX_ENDPOINT='http://<HOST_IP_ADDRESS>:4000/graphql' \
       docker.io/apache/incubator-kie-sonataflow-management-console:main

  2. In a separate terminal, start Sonataflow Dev App for the Data Index service.

      cd ../sonataflow-dev-app
      pnpm start

• There is also a pre-built image of this branch on my Quay profile:

quay.io/fabrizio_antonangeli/incubator-kie-sonataflow-management-console:kie-tools-issues-2695

[fantonangeli]

Changed this PR to "DO NOT MERGE": Doing some extra tests I found a bug in the communication between MGMT deployment and DI deployment on OCP using SSL.

[fantonangeli]

The fix is working on my OCP

[tiagobento]

Not blocking this PR, but this seems to fix a very specific problem on a very specific environment... Would it be possible to plan for some E2E tests that validate it?

[fantonangeli]

@ricardozanini now I'm trying with 172.18.0.1 , which should be the docker bridge IP and should not change between environments