ansible-role-proxmox: Getting error when creating a new cluster

@lae I did some digging through the source code as well as tried some different configurations for the command. It feels like this isn’t a bug in the playbook itself.

root@mymachine:~# pvecm add 192.168.0.xxx -use_ssh=yes -link0 192.168.0.xxx
format error at /usr/share/perl5/PVE/JSONSchema.pm line 1866 format: property is missing and it is not optional type: Invalid schema definition. description: Invalid schema definition. optional: Invalid schema definition.

Throws out an error, even running the command directly on the host.

If I run the same command with -use_ssh=no or without -use_ssh at all, it works.

In the code, there is an entire section dedicated to when the use_ssh flag is false.

	    if (!$param->{use_ssh}) {
		my $password = PVE::PTY::read_password("Please enter superuser (root) password for '$host': ");

		delete $param->{use_ssh};
		$param->{password} = $password;

		my $local_cluster_lock = "/var/lock/pvecm.lock";
		PVE::Tools::lock_file($local_cluster_lock, 10, \&PVE::Cluster::Setup::join, $param);

		if (my $err = $@) {
		    if (ref($err) eq 'PVE::APIClient::Exception' && defined($err->{code}) && $err->{code} == 501) {
			$err = "Remote side is not able to use API for Cluster join!\n" .
			       "Pass the 'use_ssh' switch or update the remote side.\n";
		    }
		    die $err;
		}
		return; # all OK, the API join endpoint successfully set us up
	    }

So the error has to happen when it’s set. I followed the code and my hunch is that it happens in this line:

The link0 flag doesn’t get used at all when use_ssh is not set. I just traced the code as much as I could without running it.

Long story short, I think it might be a bug on the parser from the PVE code.

Thanks @Moritz0708 for letting me know how to use that. I’m just now picking up Ansible. That helped me getting the params for the command.

I have the same Error. The 2 Servers not Join the Proxmox Cluster.

Its on the Task: TASK [lae.proxmox : Add node to Proxmox cluster]

Error: changed: [Proxible-2] => { "changed": true, "cmd": [ "pvecm", "add", "123.456.789.012", "-use_ssh", "-link0", "123.456.789.013" ], "delta": "0:00:01.808788", "end": "2020-05-15 08:51:08.420647", "invocation": { "module_args": { "_raw_params": "pvecm add 123.456.789.012 -use_ssh -link0 123.456.789.013 ", "_uses_shell": false, "argv": null, "chdir": null, "creates": "/etc/pve/corosync.conf", "executable": null, "removes": null, "stdin": null, "stdin_add_newline": true, "strip_empty_ends": true, "warn": true } }, "rc": 0, "start": "2020-05-15 08:51:06.611859", "stderr": "format error at /usr/share/perl5/PVE/JSONSchema.pm line 1866 format: property is missing and it is not optional description: Invalid schema definition. optional: Invalid schema definition. type: Invalid schema definition.", "stderr_lines": [ "format error at /usr/share/perl5/PVE/JSONSchema.pm line 1866 format: property is missing and it is not optional description: Invalid schema definition. optional: Invalid schema definition. type: Invalid schema definition." ], "stdout": "", "stdout_lines": [] }

@lae Sure, it’s pretty plain, mostly defaults:

pve_group: pve_familia
pve_reboot_on_kernel_update: true
pve_run_system_upgrades: true
pve_cluster_enabled: yes
pve_cluster_clustername: familia

all.yaml

ntp_manage_config: true
ntp_servers:
  - time-a-g.nist.gov iburst
  - time-a-wwv.nist.gov iburst
  - time-a-b.nist.gov iburst
  - utcnist.colorado.edu iburst
  - time-b-wwv.nist.gov	iburst