Docker compose: container is unhealthy

[emlautarom1]

After building all Docker images on a clean repo, running docker compose up results in an error after ~15 seconds: dependency failed to start: container mainnet-anvil is unhealthy

The issue seems to be due to the following healtcheck

https://github.com/NethermindEth/near-sffl/blob/61395b7fb789cf6c5e3ecce2af572926c8d6b19e/docker-compose.yml#L37

If you attach to the mainnet-anvil container logs you can see that the application is working as intended, and if you run the same health check command from your host machine (nc -zv localhost 8545) you can see that the service is reachable (this is due to the mapping 8545:8545 from the container to the host).

Running a shell inside the container shows the root of the problem: attach to the container using docker exec -it mainnet-anvil /bin/sh, run nc -zv localhost 8545 and the exit code (echo $?) should be a 1 instead of a 0.

The issue then seems to be with the command used to perform the healthcheck.

PC Specs

• OS: Fedora 39 Workstation x86_64
• CPU: Ryzen 5 3600
• Docker version: 24.0.5

[Hyodar]

Curiously enough this isn't a problem here. I can follow the same steps, but nc does have a 0 exit code. Still, 0.0.0.0 works as well, which makes sense. So, no problem with the approach in #164.