ManageIQ doesn't retreive volumes (block storages) and images from Openstack

Hello,

I have created a cloud provider in ManageIQ. It has retreived all data from openstack, except the images and the volumes.

image

image

ManageIQ version is ivanchuk-7.20200811162055_76d6567
Openstack version is Juno
It contains glance for images and cinder+cinderv2 for volumes

image

I suspect this is the same problem for images and volumes, but which problem ???

Thanks in advance,
Véro

@agrare Can you help out here?

Actually, Openstack version is Kilo (sorry for this mistake)

Hey @vlavaud do you see any errors in the /var/www/miq/vmdb/log/evm.log or fog.log?

We’re collecting Images and Volumes from two different services so not highly likely that it is the same problem.

Yes, I get errors in evm.log :
slight_smile:[----] E, [2020-10-07T10:22:59.752192 #2324:9185fc] ERROR – : MIQ(MiqQueue#deliver) Message id: [2627], Error: [Unable to obtain a collection: ‘list_backups_detailed’ in a service: ‘cinder’ through API. Please, fix your OpenStack installation and run refresh again.]
[----] E, [2020-10-07T10:22:59.752350 #2324:9185fc] ERROR – : [ManageIQ::Providers::BaseManager::Refresher::PartialRefreshError]: Unable to obtain a collection: ‘list_backups_detailed’ in a service: ‘cinder’ through API. Please, fix your OpenStack installation and run refresh again. Method:[block (2 levels) in class:LogProxy]
[----] E, [2020-10-07T10:22:59.752508 #2324:9185fc] ERROR – : /var/www/miq/vmdb/app/models/manageiq/providers/base_manager/refresher.rb:67:in `refresh’

No errors in fog.log, only warnings.

Hello ALL,
I’m working with Véronique on this issue and I’d like to add some info.

Concerning the image issue, ManageIQ sends a request to list images but Openstack says ‘NOT FOUND’ as seen in fog.log :

[----] W, [2020-10-02T10:27:39.519419 #5748:11d6608] WARN – : MIQ(OpenstackHandle::Handle.accessor_for_accessible_tenants) HTTP 404 Error during OpenStack
request. Skipping inventory item Image #<Proc:0x0000000016e94e70@/usr/local/lib/ruby/gems/2.5.0/bundler/gems/manageiq-providers-openstack-41275114e535/lib/manageiq/providers/openstack/legacy/openstack_handle/multi_tenancy/loop.rb:7 (lambda)>
Expected([200]) <=> Actual(404 Not Found)
excon.error.response
:body => "404 Not Found\n\nThe resource could not be found.\n\n "
:cookies => [
]
:headers => {
“Content-Length” => “52”
“Content-Type” => “text/plain; charset=UTF-8”
“Date” => “Fri, 02 Oct 2020 14:27:41 GMT”
“X-Openstack-Request-Id” => “req-82e94526-d324-49e9-86f8-a7632d08aefb”
}
:host => “10.166.153.105”
:local_address => “10.166.153.199”
:local_port => 54628
:path => “/v1/v2/images” <<<<<<<<<<<<<<<<<<
:port => 9292 ==================> glance
:reason_phrase => “Not Found”
:remote_ip => “10.166.153.105”
:status => 404
:status_line => “HTTP/1.1 404 Not Found\r\n”

whereas ‘openstack catalog list’ tells us /v1

publicURL: http://10.166.153.105:9292/v1

And ‘openstack image list --debug’ shows also /v1
REQ: curl -g -i -X GET http://10.166.153.105:9292/v1/images -H "User-Agent…

I don’t know why ManageIQ uses /v1/v2

Any ideas to fix that ?

Many thanks in advance

Any idea why ManageIQ adds V2 in the path ?
Maybe a problem of versions compatibility ?

Thanks,