Oracle Cloud hosting with ARM-architecture

AI-generated summary

In a discussion about Oracle Cloud hosting with ARM architecture, terraboss reports a failed Discourse installation due to a Git fetch error (exit 128), specifically Could not resolve host: github.com. Falco suggests a network or DNS issue, noting that while the installer downloaded, the actual Discourse code fetch failed. terraboss confirms the error persists despite disabling firewalls. Falco suspects Oracle’s custom distro may cause DNS issues in containers and recommends trying a standard ARM distro like Debian. terraboss clarifies they’re using Canonical Ubuntu 22.04 on an Ampere Altra 80C with 6GB RAM, but the project has since been abandoned — they plan to retry later on a Raspberry Pi. The core issue appears to be DNS resolution within the ARM environment, possibly tied to Oracle Cloud’s networking or container configuration.

I‘ve read about your experimental ARM support.

However, I am not sure if I have paid attention to everything that has to consider for ARM.

In the end, the installation was stopped. The error message seems unhelpful to me.

What do you think?
discourse/docs/INSTALL-cloud.md at main · discourse/discourse · GitHub

FAILED
--------------------
Pups::ExecError: cd /var/www/discourse && git fetch --depth 1 origin tests-passed failed with return #<Process::Status: pid 141 exit 128>
Location of failure: /usr/local/lib/ruby/gems/2.7.0/gems/pups-1.1.1/lib/pups/exec_command.rb:117:in `spawn'
exec failed with the params {"cd"=>"$home", "hook"=>"code", "cmd"=>["git reset --hard", "git clean -f", "git remote set-branches --add origin main", "git remote set-branches origin $version", "git fetch --depth 1 origin $version", "git checkout $version", "mkdir -p tmp", "chown discourse:www-data tmp", "mkdir -p tmp/pids", "mkdir -p tmp/sockets", "touch tmp/.gitkeep", "mkdir -p                    /shared/log/rails", "bash -c \"touch -a           /shared/log/rails/{production,production_errors,unicorn.stdout,unicorn.stderr,sidekiq}.log\"", "bash -c \"ln    -s           /shared/log/rails/{production,production_errors,unicorn.stdout,unicorn.stderr,sidekiq}.log $home/log\"", "bash -c \"mkdir -p           /shared/{uploads,backups}\"", "bash -c \"ln    -s           /shared/{uploads,backups} $home/public\"", "bash -c \"mkdir -p           /shared/tmp/{backups,restores}\"", "bash -c \"ln    -s           /shared/tmp/{backups,restores} $home/tmp\"", "chown -R discourse:www-data /shared/log/rails /shared/uploads /shared/backups /shared/tmp", "find public/plugins/ -maxdepth 1 -xtype l -delete"]}bootstrap failed with exit code 128
** FAILED TO BOOTSTRAP ** please scroll up and look for earlier error messages, there may be more than one.
./discourse-doctor may help diagnose the problem.

Maybe a network error? Have you tried again? Is the error consistent across multiple tries?

Yes, it happens multiple times.

Thank you for this hint. I guess, the default firewall restrictions are the reason for this. I found the key hint inside the log …

fatal: unable to access ‘GitHub - discourse/discourse: A platform for community discussion. Free, open, simple. · GitHub’: Could not resolve host: github.com

Yes, if the installer can’t access the internet it’s going to be a bad time.

Very strange, especially because Discourse seems to be successfully downloaded.

This must be something special.

You successfully downloaded the Discourse “installer”, but downloading Discourse is what would happen on the failing step.

Are you using the Oracle weird distro? Unless I’m mistaken it has some issues with DNS inside the container that requires extra step configuring the Docker service.

Maybe try running a more common arm compatible distro like debian?

  • Ampere Altra 80C, 6 GB Ram
  • Canonical Ubuntu 22.04, Image-Build:2022.08.10-0

Actually, it should have worked. I even disabled the Oracle Cloud and Ubuntu firewall completely as a test. However, without success.

The project has died in the meantime.

Will maybe try it another time with the Raspberry Pi.