Hi everyone,

I have another question about the cloud, especially the crowbar webinterface.
When I want to list my active roles in the dashboard, I get an error saying:

We're sorry, but something went wrong
Please consult your log files or double check the URL you're trying to reach. If this is an application error, we encourage you to visit the SUSE support site for more information. Return to the dashboard.
The output of production.log provides:

I, [2015-05-27T10:43:22.339245 #9636]  INFO -- : Completed 500 Internal Server Error in 493ms
F, [2015-05-27T10:43:22.341222 #9636] FATAL -- : 
ActionView::Template::Error (undefined method `[]' for nil:NilClass):
     9:       .panel.panel-default
    10:         .panel-heading
    11:           %h2
    12:             = display_name_for(barclamp)
    13:         .panel-body
    14:           - roles.map { |role| { :role => role, :elements => role.element_nodes(@roles, @nodes) } }.each do |locals|
    15:             - unless locals[:elements].empty?
  lib/barclamp_catalog.rb:67:in `display_name'
  lib/barclamp_catalog.rb:35:in `method_missing'
  app/helpers/barclamps_helper.rb:52:in `display_name_for'
  app/views/dashboard/active_roles.html.haml:12:in `block in _app_views_dashboard_active_roles_html_haml__807530730056444146_40051540'
  app/views/dashboard/active_roles.html.haml:8:in `each'
  app/views/dashboard/active_roles.html.haml:8:in `_app_views_dashboard_active_roles_html_haml__807530730056444146_40051540'
I have seen the roles before, so this is something that came up after our successful installation, and we didn't have any major changes.
What I tried was to restart rcapache2 and rccrowbar, no effect. All our nodes have been restarted multiple times since the installation due to updates and patches, but still the same error.

Our setup consists of three nodes, admin and control node are Xen VMs running on SLES11-SP3, compute node is a physical machine running SLES12. We use almost all default settings, the cloud itself is working fine. We are able to upload images and run instances. As the crowbar webinterface is running on admin node, I assume the cause is somewhere on the admin node.

Is anyone able to help or at least has other ideas where to look for hints? That would be great!