Trying to debug cypher error, or completely clean an attempted installation

I installed ON cleanly on a local machine as a test.

In attempting to script the install via Ansible for a remote host, I accidentally launched the services before changing the password.

I did a dance with sqlite3 to change the password (successfully), but ran into another error. So, I:

  • Did an apt purge
  • apt autoremove
  • Removed any lingering files in /var/lib/one, /etc/one, and anywhere else I could find things.
  • Attempted a clean install.
  • Modified /etc/one/sunstone-server.conf to use host ‘0.0.0.0’ instead of ‘127.0.0.1’

and, despite having tried to completely purge and restart the install, I consistently get an error:

ArgumentError at /

key must be 32 bytes

  • file: server_cipher_auth.rb
  • location: key=
  • line: 130

I get this error:

  • When the host is set to 127.0.0.1 and I use lynx/curl on the local machine.
  • When the host is set to 0.0.0.0 and I use lynx/curl locally.
  • When the host is 127.0.0.1 and I attempt to connect, via VPN, to the host from my machine.
  • When the host is 0.0.0.0 + local connect via VPN.

/var/log/one/sunstone.log only says:

Thu Nov 21 21:35:41 2019 [I]: 192.168.2.99 - - [21/Nov/2019:21:35:41 -0500] “GET /sinatra/500.png HTTP/1.1” 401 - 0.0044
Thu Nov 21 21:35:41 2019 [I]: 192.168.2.99 - - [21/Nov/2019:21:35:41 -0500] “GET /favicon.ico HTTP/1.1” 401 - 0.0032

and sysctl status says:

● opennebula-sunstone.service - OpenNebula Web UI Server
Loaded: loaded (/lib/systemd/system/opennebula-sunstone.service; enabled; vendor preset: enabled)
Active: active (running) since Thu 2019-11-21 21:39:11 EST; 2min 22s ago
Process: 12862 ExecStartPre=/bin/sh -c [ -s “/var/lib/one/.one/sunstone_auth” ] || /bin/sleep 2s (code=exited, status=0/SUCCESS)
Main PID: 12868 (ruby)
Tasks: 22 (limit: 4915)
CGroup: /system.slice/opennebula-sunstone.service
└─12868 /usr/bin/ruby /usr/lib/one/sunstone/sunstone-server.rb

Nov 21 21:39:20 filestore ruby[12868]: /usr/lib/ruby/vendor_ruby/sinatra/base.rb:182:in call' Nov 21 21:39:20 filestore ruby[12868]: /usr/lib/ruby/vendor_ruby/sinatra/base.rb:2013:in call’
Nov 21 21:39:20 filestore ruby[12868]: /usr/lib/ruby/vendor_ruby/sinatra/base.rb:1487:in block in call' Nov 21 21:39:20 filestore ruby[12868]: /usr/lib/ruby/vendor_ruby/sinatra/base.rb:1787:in synchronize’
Nov 21 21:39:20 filestore ruby[12868]: /usr/lib/ruby/vendor_ruby/sinatra/base.rb:1487:in call' Nov 21 21:39:20 filestore ruby[12868]: /usr/lib/ruby/vendor_ruby/thin/connection.rb:86:in block in pre_process’
Nov 21 21:39:20 filestore ruby[12868]: /usr/lib/ruby/vendor_ruby/thin/connection.rb:84:in catch' Nov 21 21:39:20 filestore ruby[12868]: /usr/lib/ruby/vendor_ruby/thin/connection.rb:84:in pre_process’
Nov 21 21:39:20 filestore ruby[12868]: /usr/lib/ruby/vendor_ruby/eventmachine.rb:1054:in call' Nov 21 21:39:20 filestore ruby[12868]: /usr/lib/ruby/vendor_ruby/eventmachine.rb:1054:in block in spawn_threadpool’

At this point, I have no idea if I have something cached somewhere from a previous install, or if this is a bug, or if I have done something awful to my Ruby install… I would be very happy to completely/cleanly uninstall ON, and do a clean install from scratch… but, that would suggest I know everything that needs to go, because something is breaking.

When I use oneuser show and oneuser list on the host, things report back cleanly.

Thoughts? Any ideas welcome.

I’m starting to believe that I failed to add the correct “apt” repository… I just noticed that, at the bottom of the login page, it says I’m running 4.12…

I’ll close this thread when I discover the full depths of my own mistakes…

/me embarrassed coughing…

My Ansible automation must have failed to add the apt package repository correctly. As a result, I installed an old version of OpenNebula. This was not immediately obvious to me. A big clue was that the Ruby was using deprecated functions… which didn’t happen on my first install.

I successfully removed the old OpenNebula, made a bit of a mess of my Ruby installation, and have managed to get it running. I kinda wish I had been working in a VM, so that my Ruby installation wouldn’t have suffered quite so much, but that’s my own fault.

I have ON running, and all of this thread can be considered closed. Perhaps it will be useful to someone else who makes the same mistake I did: look at the fine print at the bottom of the webpage. Know what version you’re running.

Many thanks!