Bug 830779 - 3.1 - rhev-h: host install is initiated with localhost.localhost instead of name configured in console
3.1 - rhev-h: host install is initiated with localhost.localhost instead of n...
Product: Red Hat Enterprise Virtualization Manager
Classification: Red Hat
Component: ovirt-engine (Show other bugs)
x86_64 Linux
medium Severity medium
: ---
: ---
Assigned To: Doron Fediuck
Pavel Stehlik
Depends On:
  Show dependency treegraph
Reported: 2012-06-11 08:01 EDT by Dafna Ron
Modified: 2016-04-26 13:56 EDT (History)
13 users (show)

See Also:
Fixed In Version:
Doc Type: Bug Fix
Doc Text:
Story Points: ---
Clone Of:
Last Closed: 2012-08-07 11:37:30 EDT
Type: Bug
Regression: ---
Mount Type: ---
Documentation: ---
Verified Versions:
Category: ---
oVirt Team: Infra
RHEL 7.3 requirements from Atomic Host:
Cloudforms Team: ---

Attachments (Terms of Use)
log and screen shot (107.24 KB, application/x-gzip)
2012-06-11 10:30 EDT, Dafna Ron
no flags Details

  None (edit)
Description Dafna Ron 2012-06-11 08:01:14 EDT
Description of problem:

I added the rhev-m in the rhev-h console and the install did not automatically intiate (as in the host was not added to my setup with waiting for approve). 
only when I added the host from the UI the host was added with waiting for approve 
tested for both 3.0 and 3.1 rhe-v setups. 

Version-Release number of selected component (if applicable):

Red Hat Enterprise Virtualization Hypervisor release 6.3 (20120606.3)

How reproducible:


Steps to Reproduce:
1. install rhev-h
2. try to initiate the install from the console 
Actual results:

host is not added to the rhevm setup. 

Expected results:

host should be added to the rhevm setup

Additional info:
Comment 1 Perry Myers 2012-06-11 08:42:05 EDT
This functionality is controlled by vdsm (RHEV-M registration) so moving to the correct component.
Comment 2 Mike Burns 2012-06-11 09:27:06 EDT
first thing that vdsm team will ask for are the vdsm-reg logs from rhev-h.
Comment 3 Dafna Ron 2012-06-11 10:23:31 EDT
I did some more testing, I found that we do intiate the registration but it takes time because we are failing on https. 
having said that, I am not closing this bug only altering it since the registration is done with localhost.locathost although the network is configured.

MainThread::DEBUG::2012-06-11 14:25:08,351::vdsm-reg-setup::131::root::registerVDS URI= /RHEVManagerWeb/VdsAutoRegistration.aspx?vds_ip=10.35.XXX.15&vds_name=localhost.localdomain&vds_unique_id=44454C4C-3800-1046-8035-B1C04F34344A_00%3A22%3A19%3A14%3A7e%3Aca&port=54321&__VIEWSTATE=

and I am adding logs and screenshots (thanks mike :) forgot to add).
Comment 4 Dafna Ron 2012-06-11 10:30:26 EDT
Created attachment 590950 [details]
log and screen shot
Comment 5 Dafna Ron 2012-06-11 10:37:58 EDT
I am lowering priority since once we approve the host we can change the name manually.
but I am deducting that if I added more then one host we will fail on next host's with CanDoAction 'host name already exists'.
Comment 6 Itamar Heim 2012-06-11 16:35:05 EDT
there's supposed to be logic that sends the ip if hostname is localhost rather than configured (not sure if in engine or vdsm though)
1. need to check why localhost was kept.
2. need to check why localhost was used to begin with, rather than the configured name
Comment 7 Doron Fediuck 2012-07-10 11:12:34 EDT
This is actually working by design.
In registration the backend will not change the machine's host name, only report
it, so it will be used by the backend.
The backend needs to be able to handle name duplication, for cases like this one
or dhacp names (dhcp-1-173 as a host name). So there's a mechanism to rename the
previous host name with the duplicate name, assuming that the new one is the correct one. 
You will not get a can-do-action error, since we're testing for machine ID, and not name.

Please try to register 2 localhost.localdomain rhev-h machines, and report the results.
Comment 16 Doron Fediuck 2012-08-07 11:37:30 EDT
After some internal checks, with Pavel's good assistance, we've established 
that this is working by design, as stated in comment 7.

Closing as this is not a bug.

Note You need to log in before you can comment on or make changes to this bug.