Just docker configuration

[bitchecker]

Describe the bug
Hi,
will be great to have some documentation on how to run just the application instead of the entire stack as container.

Maybe for k8s deployment or simple "docker server" with dedicated databases already running.

[Clivern]

I think something like this should work

$ docker run -d \
  --name lynx \
  --restart unless-stopped \
  -p 4000:4000 \
  -e APP_NAME=Lynx \
  -e APP_PORT=4000 \
  -e APP_SECRET=koPmu7TJCwD8mttV9vgWUeU7iuu/zTPOR3sX4UalM9KkYEVGPfyi0PeTVzu1TT8C \
  -e APP_HOST=localhost \
  -e APP_HTTP_SCHEMA=http \
  -e DB_USERNAME=doadmin \
  -e DB_PASSWORD=AVNS_QYES3UaJRovCIfBOB93 \
  -e DB_HOSTNAME=db-postgresql-ams3-41583-do-user-1532625-0.e.db.ondigitalocean.com \
  -e DB_DATABASE=lexi \
  -e DB_PORT=25060 \
  -e DB_SSL=on \
  -e DB_CA_CERTFILE_PATH=/app/ca-certificate.crt \
  -e MIX_ENV=prod \
  -v /root/ca-certificate.crt:/app/ca-certificate.crt \
  clivern/lynx:0.12.7 sh -c "/app/bin/migrate && /app/bin/server"

[bitchecker]

Hi @Clivern,
thanks for reply, and sorry for delay!

The options that you provided works, but what I found is "strange":

• Container needs double ram compared with direct build
• I can see that the migration phase works (so I can see tables on database) but If I try to reach the application I get "Internal Server Error" because logs are reporting timeout on database.

It's really strange, this behaviour. If it's able to create tables and everything is needed, why after that I get timeouts? 🤔

[Clivern]

Hi @Clivern, thanks for reply, and sorry for delay!

The options that you provided works, but what I found is "strange":

• Container needs double ram compared with direct build
• I can see that the migration phase works (so I can see tables on database) but If I try to reach the application I get "Internal Server Error" because logs are reporting timeout on database.

It's really strange, this behaviour. If it's able to create tables and everything is needed, why after that I get timeouts? 🤔

In my opinion, it should work in all cases. I tested it with local postgresql, and dockerized db and managed postgresql. they all work.

Anyways if you see these issues, please share the log records. it can help to identify the root cause.