Description of problem: Missing Guest OS in dashboard reports in Openstack 8/10 Version-Release number of selected component (if applicable): 5.8.2.3 How reproducible: 100% Actual results: No data shown Expected results: Should show Guest OS
Marek: for dashboard to work correctly, I think we need what azure does: https://github.com/Ladas/manageiq-providers-azure/blob/0991914c6e08e60122ed49a8684b4b84772d229b/app/models/manageiq/providers/azure/cloud_manager/refresh_parser.rb#L241 and https://github.com/Ladas/manageiq-providers-azure/blob/0991914c6e08e60122ed49a8684b4b84772d229b/app/models/manageiq/providers/azure/cloud_manager/refresh_parser.rb#L241
I've done the same fix for the AWS provider, so follow that https://github.com/ManageIQ/manageiq-providers-amazon/pull/376
It looks like there are possibly two different things at work here. I'm not sure which is the primary one driving this BZ: 1) The Guest OS is showing up as "Unknown" on the dashboard, although the Operating System you see when going into the node show/details page is actually "linux". The dashboard view should be updated to show the same as the node details (it appears that while the OS is listed as "linux" in the node/host ActiveRecord object, the MiqReport that the dashboard widget pulls from does not list the operating system properly) 2) The Operating System field for OpenStack infra nodes is hard-coded to "linux" in the refresh parser rather than actually getting the information from OpenStack. In this case, we need to see if that information is available via the APIs we're using to refresh and, if so, incorporate it into the AR object updates. I'm wondering if 1) and 2) are really separate bugs. 2) would depend on 1). It could be that 1) would have a higher PM score than 2), which would make the more important fix easier than the less important one -- if that's the case, we'd probably want two BZs here. It could be the reverse, in which case 1) isn't particularly useful on its own as a fix. In that case, we'd probably want a single BZ (this one) and incorporate both changes into the same fix.
In any case, I'm just seeing the comment from Ladislav now (I think I must have refreshed this page right before he submitted it earlier today). I'll start with that.
https://github.com/ManageIQ/manageiq-providers-openstack/pull/193
New commit detected on ManageIQ/manageiq-providers-openstack/master: https://github.com/ManageIQ/manageiq-providers-openstack/commit/4ee191f1c40d9d3f63c090cd00f3a050a40649d9 commit 4ee191f1c40d9d3f63c090cd00f3a050a40649d9 Author: Scott Seago <sseago> AuthorDate: Tue Jan 9 09:03:49 2018 -0500 Commit: Scott Seago <sseago> CommitDate: Fri Jan 12 10:34:57 2018 -0500 BZ 1520565: Track guest OS for openstack images and VMs https://bugzilla.redhat.com/show_bug.cgi?id=1520565 Use os_distro and os_version properties on glance images to track operating system information on templates and VMs. .../providers/openstack/cloud_manager/refresh_parser.rb | 13 +++++++++++++ .../openstack/inventory/parser/cloud_manager.rb | 16 ++++++++++++++++ .../openstack/inventory/persister/cloud_manager.rb | 1 + .../openstack/inventory/persister/target_collection.rb | 4 ++-- .../inventory_collection_default/cloud_manager.rb | 14 ++++++++++++++ .../providers/openstack/refresh_parser_common/images.rb | 7 +++++++ .../openstack/cloud_manager/refresh_spec_common.rb | 10 +++++----- .../openstack/infra_manager/refresher_rhos_juno_spec.rb | 2 +- 8 files changed, 59 insertions(+), 8 deletions(-)
(forgot to remove needinfo when setting it to POST)
Verified on 5.10.0.2