[Bug]: ERROR in instillation when update yesterday

[dhunganasagar]

Contact Details

dhunganasagar@yahoo.com

What happened?

Error response from daemon: driver failed programming external connectivity on endpoint caddy (cef7de5b0883ca427c4b6d01a7687fb9c9e2b76316914e2395856ed893c4e027): Error starting userland proxy: listen tcp4 144.24.97.253:443: bind: cannot assign requested address

Version

v0.30.0

What OS are you using?

Linux

Relevant log output

No response

Contributing guidelines

• Yes, I did.

[abhishek9686]

what steps did you follow to upgrade?

[dhunganasagar]

what steps did you follow to upgrade?

I try to Install fresh, all firewall ports are open , all step complete, (https://docs.netmaker.io/docs/server-installation/quick-install#quick-install-script) but give error

[abhishek9686]

can you try a fresh install?
we merged the release changes into master

[dhunganasagar]

fresh instillation
ubuntu 24.4
give error

Error response from daemon: driver failed programming external connectivity on endpoint caddy (2953e7530423ee0209104253d8c128c629dbb9c7ca0a407b396e47e76578aa8f): failed to bind port 19:80/tcp: Error starting userland proxy: listen tcp4 19:80: bind: cannot assign requested address

just install

[abhishek9686]

can you share the output of this command sudo lsof -i -P -n

[cocoonkid]

make netmaker
depend_on: caddy
in your docker-compose.yml.
that way it will start only when caddy is already running and has taken over 0.0.0.0.0:443.
netmaker will then switch to the old default port.
That should make your setup work.

like this. I think.

 caddy:
    healthcheck:
      test: ["CMD", "curl", "-f", "https://localhost:443"]
      interval: 10s
      timeout: 5s
      retries: 3

  netmaker:
    depends_on:
      caddy:
        condition: service_healthy

[abhishek9686]

make netmaker depend_on: caddy in your docker-compose.yml. that way it will start only when caddy is already running and has taken over 0.0.0.0.0:443. netmaker will then switch to the old default port. That should make your setup work.

like this. I think.

 caddy:
    healthcheck:
      test: ["CMD", "curl", "-f", "https://localhost:443"]
      interval: 10s
      timeout: 5s
      retries: 3

  netmaker:
    depends_on:
      caddy:
        condition: service_healthy

this is not relevant

[cocoonkid]

it's absolutely relevant. Alright . You do you.

[arpadgabor]

The issue stems from the updated docker-compose in 8546f85. Removing $SERVER_HOST: will make it work, but rerunning the script overwrites the compose file so you must edit the script to not replace the compose file.