Can't launch after update

[ac2323]

I updated my Docker Docuseal image recently and noticed that it doesn't start. I'm on an Ubuntu LXC 22.04.4 Below is the error log:

WARNING: Nokogiri was built against libxml version 2.11.6, but has dynamically loaded 2.11.7
=> Booting Puma
=> Rails 7.1.2 application starting in production 
=> Run `bin/rails server --help` for more startup options
Exiting
/usr/local/bundle/gems/rack-2.2.8/lib/rack/server.rb:446:in `delete': Resource busy @ apply2files - /dev/null (Errno::EBUSY)
        from /usr/local/bundle/gems/rack-2.2.8/lib/rack/server.rb:446:in `check_pid!'
        from /usr/local/bundle/gems/rack-2.2.8/lib/rack/server.rb:307:in `start'
        from /usr/local/bundle/gems/railties-7.1.2/lib/rails/commands/server/server_command.rb:38:in `start'
        from /usr/local/bundle/gems/railties-7.1.2/lib/rails/commands/server/server_command.rb:145:in `block in perform'
        from <internal:kernel>:90:in `tap'
        from /usr/local/bundle/gems/railties-7.1.2/lib/rails/commands/server/server_command.rb:136:in `perform'
        from /usr/local/bundle/gems/thor-1.3.0/lib/thor/command.rb:28:in `run'
        from /usr/local/bundle/gems/thor-1.3.0/lib/thor/invocation.rb:127:in `invoke_command'
        from /usr/local/bundle/gems/railties-7.1.2/lib/rails/command/base.rb:178:in `invoke_command'
        from /usr/local/bundle/gems/thor-1.3.0/lib/thor.rb:527:in `dispatch'
        from /usr/local/bundle/gems/railties-7.1.2/lib/rails/command/base.rb:73:in `perform'
        from /usr/local/bundle/gems/railties-7.1.2/lib/rails/command.rb:71:in `block in invoke'
        from /usr/local/bundle/gems/railties-7.1.2/lib/rails/command.rb:149:in `with_argv'
        from /usr/local/bundle/gems/railties-7.1.2/lib/rails/command.rb:69:in `invoke'
        from /usr/local/bundle/gems/railties-7.1.2/lib/rails/commands.rb:18:in `<main>'
        from <internal:/usr/local/lib/ruby/3.2.0/rubygems/core_ext/kernel_require.rb>:37:in `require'
        from <internal:/usr/local/lib/ruby/3.2.0/rubygems/core_ext/kernel_require.rb>:37:in `require'
        from /usr/local/bundle/gems/bootsnap-1.17.0/lib/bootsnap/load_path_cache/core_ext/kernel_require.rb:32:in `require'
        from /app/bin/rails:4:in `<main>'

Any ideas?

[omohokcoj]

@ac2323 what's your docker version?

[ac2323]

@ac2323 what's your docker version?

Docker version 25.0.3, build 4debf41

[ac2323]

Noticed this error as well: A server is already running. Check /app/tmp/pids/server.pid

I was able to resolve this by searching for that server.pid file and removing all copies of it. Docuseal was then able to launch correctly.

[ac2323]

Reopening since the issue persists after a reboot. The server.pid file gets recreated and Docuseal will not start until those files are removed. Furthermore, updating the docker image yields the same error as my original post and there is no server.pid file to delete.

[omohokcoj]

@ac2323 which version of docuseal docker image are you using, also on what is your docker host OS?

[ac2323]

@ac2323 which version of docuseal docker image are you using, also on what is your docker host OS?

Using the latest (1.4.5). OS is Ubuntu 22.04.4 LXC

[omohokcoj]

@ac2323 looks like /dev/null is not available on LXC - but it should work with docker

[ac2323]

I have other containers working just fine.

I was able to restore to a previous backup running Docuseal 1.4.2 which requires deleting those 2 server.pid files.

[omohokcoj]

In the latest 1.4.5 we use /dev/null as server pid file - and looks like it's causing issues in LXC (/dev/null is not available on LXC). Afaik there are no easy ways to disable PID file - /dev/null was the only solution @ac2323

[ac2323]

Confirmed that this is limited to 1.4.5 as I was able to launch 1.4.4 just fine

[omohokcoj]

@ac2323 to fix it on LXC and an 1.4.5 version use -e PIDFILE=/tmp/server.pid in order to write pid file into tmp folder instead of /dev/null

[econcepcion]

Can you point me in the right direction? How do I accomplish this?

use -e PIDFILE=/tmp/server.pid in order to write pid file into tmp folder instead of /dev/null

[omohokcoj]

@econcepcion yep, correct

[econcepcion]

I am lost. Do I add this to the compose file?

[omohokcoj]

@econcepcion yep: - PIDFILE=/tmp/server.pid

[econcepcion]

I added this... - PIDFILE=/tmp/server.pid

I am sure I have not added it correctly, cause it is not working.

[omohokcoj]

@econcepcion did you try --force-recreate

[econcepcion]

Just tried --force-recreate but still getting this error.

ERROR: for app Container "59b55cdd4dd5" is unhealthy. ERROR: Encountered errors while bringing up the project.

[omohokcoj]

@econcepcion looks like now that's a new error that is not related to /dev/null

[econcepcion]

oh ok. Thanks for the help. You are a Legend.