Oops, we're sorry but something went wrong undefined method `cluster' for #<Fog::Libvirt::Compute::Server:0x0000556ccffaf870>

[fthyes]

Installed foreman_netbox but when I try to create a virtual host we receive the following error. Looks like a incompatibility to fog.... previous errors were about memory-mb and size which could be workaround by adjusting the code. But now I stuck, any ideas how to solve that?

foreman-rake errors:fetch_log request_id=bae5ef46

Warning: Logging layout is not multiline_request_pattern.
This output of this command can be incomplete.
2024-03-12T07:30:22 [I|app|bae5ef46] Started POST "/hosts" for 10.0.250.106 at 2024-03-12 07:30:22 +0000
2024-03-12T07:30:22 [I|app|bae5ef46] Processing by HostsController#create as /
2024-03-12T07:30:22 [I|app|bae5ef46] Parameters: {"utf8"=>"�", "authenticity_token"=>"UvwzsGPA7ZMMoEo/WwrTuqnknJHVcK/ZDbBiXkWSs7DcvqCRL1aAxXkTmMoceytyBvs2d+Z9l3o1NyEyAr/jiA==", "host"=>{"name"=>"adam-larish", "organization_id"=>"1", "location_id"=>"2", "hostgroup_id"=>"39", "comput
e_resource_id"=>"54", "puppet_proxy_id"=>"15", "puppet_ca_proxy_id"=>"14", "puppetclass_ids"=>[""], "managed"=>"true", "progress_report_id"=>"[FILTERED]", "type"=>"Host::Managed", "interfaces_attributes"=>{"0"=>{"destroy"=>"0", "type"=>"Nic::Managed", "identifier"=>"", "name"=>"adam
-larish", "domain_id"=>"15", "subnet_id"=>"9", "ip"=>"10.0.250.10", "ip6"=>"", "managed"=>"1", "primary"=>"1", "provision"=>"1", "virtual"=>"0", "tag"=>"", "attached_to"=>"", "compute_attributes"=>{"type"=>"bridge", "bridge"=>"xenbr250", "model"=>"virtio"}}}, "compute_attributes"=>{"
cpus"=>"1", "cpu_mode"=>"default", "memory"=>"2147483648", "start"=>"1", "volumes_attributes"=>{"0"=>{"delete"=>"", "pool_name"=>"default", "capacity"=>"10G", "allocation"=>"0G", "format_type"=>"raw"}}}, "architecture_id"=>"1", "operatingsystem_id"=>"55", "provision_method"=>"build"
, "build"=>"1", "medium_id"=>"20", "ptable_id"=>"109", "pxe_loader"=>"PXELinux BIOS", "disk"=>"", "root_pass"=>"[FILTERED]", "is_owned_by"=>"7-Users", "enabled"=>"1", "model_id"=>"", "comment"=>"", "overwrite"=>"false"}}
2024-03-12T07:30:22 [W|api|bae5ef46] param host[puppetclass*] has been deprecated in favor of host[puppet_attributes][puppetclass*]
2024-03-12T07:30:23 [W|app|bae5ef46] Scoped order is ignored, it's forced to be batch order.
2024-03-12T07:30:24 [W|app|bae5ef46] undefined method cluster' for #<Fog::Libvirt::Compute::Server:0x0000556ccffaf870> 2024-03-12T07:30:24 [I|app|bae5ef46] Backtrace for 'undefined method cluster' for #Fog::Libvirt::Compute::Server:0x0000556ccffaf870' error (NoMethodError): undefined method `cluster' for #Fog::Libvirt::Compute::Server:0x0000556ccffaf870
2024-03-12T07:30:24 [I|app|bae5ef46] Rendering common/500.html.erb
2024-03-12T07:30:24 [I|app|bae5ef46] Rendered common/500.html.erb (Duration: 8.7ms | Allocations: 1063)
2024-03-12T07:30:24 [I|app|bae5ef46] Completed 500 Internal Server Error in 1189ms (Views: 13.5ms | ActiveRecord: 179.1ms | Allocations: 88709)

Foreman version: 3.2.1
Plugins:

• foreman-tasks 6.0.1
• foreman_memcache 0.1.1
• foreman_monitoring 2.1.0
• foreman_netbox 1.1.1
• foreman_puppet 3.0.7

[optiz0r]

It looks like the ESXi and Proxmox ComputeResource objects have different methods. I had Netbox plugin working fine with ESXi compute resources, but then I moved them over to Proxmox after the Broadcom licensing shenanigans, and ran into these issues. Like the OP I renamed memorysize-mb->memory and size_gb->size. Setting cluster to nil (because I couldn't easily work out what the value should be) allows foreman to at least create the host, although it does not then end up being sync'd into netbox.

    def netbox_compute_attributes
      @netbox_compute_attributes ||= host.compute_object.yield_self do |compute_object|
        {
          vcpus: compute_object&.cpus,
          memory: compute_object&.memory,
          disk: compute_object&.volumes&.map(&:size)&.reduce(&:+),
          cluster: nil,
        }.compact
      end
    end