Fix: Docker Setup #980
Fix: Docker Setup #980
Conversation
danielmbrasil
force-pushed
the
fix/docker-setup
branch
from
76ba8f4 to
aa6872f
Compare
danielmbrasil
force-pushed
the
fix/docker-setup
branch
from
aa6872f to
978487f
Compare
| worker: bundle exec sidekiq -C config/sidekiq.yml | ||
| vite: bin/vite dev |
There was a problem hiding this comment.
There was a problem hiding this comment.
There's this issue related to the translations stuff. We can investigate later. In production this problem doesn't exist. Just in development.
There was a problem hiding this comment.
The production Procfile does not start a new process for Vite. So, I believe that by removing this process that issue has been fixed.
| worker: bundle exec sidekiq -C config/sidekiq.yml | ||
| vite: bin/vite dev | ||
| pusherfake: pusher-fake -i ${PUSHER_APP_ID:-123456} --socket-host 0.0.0.0 --socket-port ${PUSHER_WS_PORT:-45449} --web-host 0.0.0.0 --web-port ${PUSHER_PORT:-8888} -k ${PUSHER_APP_KEY:-centralakey} -s ${PUSHER_APP_SECRET:-centralsecret} | ||
| pusherfake: PUSHER_FAKE=1 pusher-fake -i ${PUSHER_APP_ID:-123456} --socket-host 0.0.0.0 --socket-port ${PUSHER_WS_PORT:-45449} --web-host 0.0.0.0 --web-port ${PUSHER_PORT:-8888} -k ${PUSHER_APP_KEY:-centralakey} -s ${PUSHER_APP_SECRET:-centralsecret} |
There was a problem hiding this comment.
There is an annoying issue with this process that looks to be a bug in Foreman. Foreman sends a SIGTERM to all child processes and then they are terminated. However, for pusher-fake specifically, the resources are not fully released by the process and it keeps the port 8888 busy. I couldn't figure out how to solve this, if you have any suggestion I'd appreciate it. For now, you can manually kill the process after exiting Foreman.
There was a problem hiding this comment.
Created this issue to investigate later.
| nginx-proxy: | ||
| image: jwilder/nginx-proxy | ||
| ports: | ||
| - "80:80" | ||
| volumes: | ||
| - /var/run/docker.sock:/tmp/docker.sock:ro |
There was a problem hiding this comment.
Since the application is no longer running on a Docker container, we can no longer bind the Nginx proxy to the application web server, so removing this service.
| adminer: | ||
| image: adminer:4.8.1 | ||
| environment: | ||
| - VIRTUAL_HOST=adminer.cm42-central.localhost | ||
| ports: | ||
| - 8080:8080 |
There was a problem hiding this comment.
Is this still needed?
There was a problem hiding this comment.
We can leave this for now. Just admin front-end to handle database stuff. We might remove this in the future.
There was a problem hiding this comment.
It always makes me happy to see your PR's @danielmbrasil 😄
I used the changes from this pr to help @julialbq with her setup 👍
Overall, LGTM 
If you want to, we can join forces to investigate this pusher-fake later on 🤙
|
We noticed there were e2e tests that haven't worked. Opened this issue |
|
Besides, opened this issue to enhance Cypress E2E cover |
What this PR do ?
Updates the Docker setup. The application no longer is run inside a Docker container. Now, it's possible to run the application infrastructure (Redis, Postgres, Memcached) in containers, while running the application itself on your machine.
Related Issues
Fixes #930.
Closes #873 (deprecated)
Screenshots (if applicable)
Additional Notes (if any)