Those log messages are the polling to see if a new model has been marked for production. It looks like there’s an issue with your token. Confirm that the .env has RASA_X_TOKEN set to the value you are seeing in the log and try a docker-compose restart.
If you do a docker-compose ps you’ll find that port 5002 is not exposed. You can create a docker-compose.override.yml to change this but the default config exposes Rasa X and the Rasa OSS api end point via the nginx container which is on port 80.
I just checked my .env file and it does indeed contain a rasa-x token variable: RASA_X_TOKEN=1dOJeqQ2RMS9JNHXLhnFiQ==
The token in the logs is 1dOJeqQ2RMS9JNHXLhnFiQ%3D%3D and this one is 1dOJeqQ2RMS9JNHXLhnFiQ==. I guess it’s not feeling too happy about ==. I changed it to a string containing only numbers. Now the error changes from 404 to 503:
[2021-02-21 08:28:44 +0000] - (sanic.access)[INFO][172.25.0.8:44590]: GET http://rasa-x:5002/api/config?token=1234567891011121314 503 40
Here are my containers and it looks like 5002 is exposed.
rasa-x_app_1 ./entrypoint.sh run python ... Up 5055/tcp
rasa-x_db-migration_1 python -m rasax.community. ... Up (health: starting) 8000/tcp
rasa-x_db_1 /opt/bitnami/scripts/postg ... Up 5432/tcp
rasa-x_duckling_1 duckling-example-exe --no- ... Up 8000/tcp
rasa-x_nginx_1 /opt/bitnami/entrypoint.sh ... Up 8080/tcp, 0.0.0.0:443->8443/tcp, 0.0.0.0:90->9090/tcp
rasa-x_rabbit_1 /opt/bitnami/scripts/rabbi ... Up 15671/tcp, 15672/tcp, 25672/tcp, 4369/tcp, 5671/tcp, 5672/tcp
rasa-x_rasa-production_1 rasa x --no-prompt --produ ... Up 5005/tcp
rasa-x_rasa-worker_1 rasa x --no-prompt --produ ... Up 5005/tcp
rasa-x_rasa-x_1 /tini -g -- sh -c user_id= ... Up 5002/tcp
rasa-x_redis_1 /opt/bitnami/scripts/redis ... Up 6379/tcp
I changed my nginx to port 90 because I had things I didn’t want to mess with on port 80. Is that a problem?