I wanted to run my own Lemmy instance but I don’t have the physical space / server to run anything massive so I was trying to use a pi 4. Using the Ansible instructions I get most of the way through up to it trying to start the docker compose instance it mentions there’s no Arm in the manifest file. I know it’s not the strongest device but does Lemmy not support the pi outright?

  • hitagi (ani.social)@ani.social
    link
    fedilink
    English
    arrow-up
    2
    ·
    1 year ago

    With the other script How does it handle setting a domain name, or the certbot / lets encrypt stuff I’m using cloudflare / a cloudflare tunnel so that’s sort of a pain.

    It only builds the ARM Docker image and publishes it to a repository which you can use in your docker-compose.yml. It doesn’t do what the Ansible playbook does where it sets up everything for you. If you want a script that sets up everything for you, check out ubergeek77’s Lemmy-Easy-Deploy.

    So my other question is, with ansible how do I find logs?

    I’m not sure what ansible names the containers but for me, I can view the logs with sudo docker logs -f lemmy_lemmy_1.

    Could it be related to nginx and certbot not communicating somehow even though the page loaded? I sorta had to modify the certbot command to use the cloudflare plugin instead of nginx if that’s an issue

    I haven’t used Cloudflare Tunnels in a long time so I’m not entirely sure. I know you can locally access Lemmy via port 80 (http://localhost:80) to check if Lemmy is working and if the black page is an nginx/cloudflared issue.

    • Weirdbeardgame@lemmy.mlOP
      link
      fedilink
      English
      arrow-up
      1
      ·
      edit-2
      1 year ago

      Apparently, the docker compose file just forces an amd build which caused a mismatch and an infinite reboot loop

      I’ve been trying the easy deploy script because it has cloudflare support. ```___ERROR [lemmy builder 6/7] RUN cargo build --release`