+++ This bug is a RHEV-M zstream clone. The original bug is: +++ +++ https://bugzilla.redhat.com/show_bug.cgi?id=1006659. +++ ====================================================================== ---------------------------------------------------------------------- Following comment by byount on September 11 at 04:25:22, 2013 Description of problem: When a VM is spun up as part of a "prestarted pool", it does not properly use the appropriate sysprep file and thus the VM does not join the Active Directory domain. Version-Release number of selected component (if applicable): 3.2.2 Steps to Reproduce: 1. Create a pool 2. Select a Windows 7 template 3. Enable a number of prestarted virtual machines Actual results: The resulting VMs are not sysprepped Expected results: The VM should sysprep properly ---------------------------------------------------------------------- Following comment by byount on September 11 at 04:27:27, 2013 This was reported by the FBI/CJIS Account # 1354910 Strategic ---------------------------------------------------------------------- Following comment by byount on September 12 at 01:49:42, 2013 I tested this on my own environment which isn't connected to an AD domain but the results of the test should still be the same. I am merely observing if the A:\sysprep.inf file is presented to the guest when I log in: 1. created a pool of 5 Windows 7 VMs 2. set the prestart number to 2 3. logged into the User Portal 4. Checked out a VM from that pool 5. Connected to the guest 6. Filled out the Windows Setup questions that you see when there is no sysprep file present 7. The VM did not reboot; it logged me in automatically 8. No A:\sysprep.inf file exists Are the customer and/or myself doing something wrong here? ---------------------------------------------------------------------- Following comment by iheim on September 16 at 05:55:37, 2013 when setting needinfo, please always set it on someone specific ---------------------------------------------------------------------- Following comment by tjelinek on September 16 at 08:44:32, 2013 This works on upstream. It has been fixed by 37952bef3fdafa6796b402a4c9a2313a0904f1f8 but it is partially a refactor and also strongly depends on the cloud init integration so there is no simple way to backport it. To fix this on the 3.2.2 it would require a downstream only patch (most probably a one line fix in VmPoolMonitor). ---------------------------------------------------------------------- Following comment by michal.skrivanek on September 16 at 09:00:37, 2013 closing in 3.3 as fixed ---------------------------------------------------------------------- Following comment by jduncan on September 16 at 13:02:01, 2013 Michael, Looking for some clarification here. The 3.2z flag was set, and there was talk of a one-line backport into 3.2. Is that possible? Is it planned? Thanks. ---------------------------------------------------------------------- Following comment by michal.skrivanek on September 16 at 13:41:24, 2013 yeah, once QA and PM acks the bug(this one) it will be cloned & fixed I can't clone it without PM and QA accepting the proposal, though. ---------------------------------------------------------------------- Following comment by jduncan on September 16 at 13:46:12, 2013 Is there any way we can get the PM/QA acks so this can move forward? This issue is affecting a large production cluster at a large public sector customer (CJIS, mentioned above), who use RHEV for their VDI solution. Their environment was just upgraded to 3.2, and the next upgrade window is a long ways off. If we could get this into 3.2z it would be very helpful. Thanks, Jamie Duncan ---------------------------------------------------------------------- Following comment by michal.skrivanek on September 16 at 15:57:45, 2013 should be tested... ---------------------------------------------------------------------- Following comment by byount on September 17 at 02:13:43, 2013 Michal, do you happen to know when this functionality was broken? Was it in the 3.2 branch? I know my former customer, State Street, loves the prestarted pool functionality in 3.1 and we're about to do a major upgrade to 3.2. If this feature is broken in 3.2, they will want to know prior to upgrade. ---------------------------------------------------------------------- Following comment by michal.skrivanek on September 17 at 10:22:48, 2013 Likely it is indeed broken in current latest 3.2.z, the request for 3.2.z fix is valid
IS there an ETA for the inclusion in the 3.2.z?
(In reply to wiley crider from comment #1) > IS there an ETA for the inclusion in the 3.2.z? please always set needinfo on someone specific. the bug is modified with a target release of 3.2.4. for anything earlier, please discuss with GSS wrt a hotfix.
checked on sf21.1, by the following flow from original 3.3 bug 1006659: 1) Added ldap server and added user from server to rhevm 2) Created vm with windows 7(operating system windows 7 64-bit) on domain of ldap server 3) Vm sealed and created template from vm 4) Created vms pool with 5 vms from template 5) Set number of prerestarted vms to two 6) Wait until os is up on prerestarted vm 7) No welcome screen, see login screen in ldap domain 8) Entered to OS with user and password from that added to rhevm from ldap domain 9) Also floppy disk with sysprep.inf exist There was some problem with the sysprep, for which the prestarted vms where not installed with windows automatically, but I checked vdsm.log, and sew the sysprep.inf file sent in vmcreate command (VM run)
Since the problem described in this bug report should be resolved in a recent advisory, it has been closed with a resolution of ERRATA. For information on the advisory, and where to find the updated files, follow the link below. If the solution does not work for you, open a new bug report. http://rhn.redhat.com/errata/RHBA-2013-1468.html