Page MenuHomeSoftware Heritage
Paste P1238

"vagrant up --no-destroy-on-error pergamon" workaround for now
ActivePublic

Authored by ardumont on Dec 7 2021, 10:27 AM.
Somehow for now, the pergamon manifest is not complete/incorrect (and it's not clear where exactly we fall short).
At boot time, we have the puppet master running while it should not. Puppet master must be served through apache (ssl configured and what not).
So we need to make the provisioning step pass twice (for some unknown reason as of yet as well).
Then connect to the node, stop and deactivate the puppet master service.
Then restart the apache service so it serves properly puppet (as master).
We may have to restart the puppetdb service as well.
```
# up and let it follow through
vagrant up --no-destroy-on-error pergamon
# force a second provisioning
vagrant provision pergamon
# connect to it
vagrant ssh pergamon
# within it, sudo
vagrant@pergamon $ sudo -i
# stop puppet-master which should not run
root@pergamon # systemctl stop puppet-master; systemctl disable puppet-master
# puppet is served through apache so let it start
root@pergamon # systemctl restart apache2; systemctl restart puppetdb
# Now, let the agent run for itself and it must be ok
root@pergamon # puppet agent --test
```
And now, from other nodes, the puppet agent should be able to run normally...

Event Timeline

ardumont edited the content of this paste. (Show Details)