Re: [ovirt-users] self hosted setup with answer file

From: "Greg Swift" <gregswift(a)gmail.com&gt; To: "Yedidyah Bar David" <didi(a)redhat.com&gt; Cc: users(a)ovirt.org Sent: Monday, February 16, 2015 5:53:12 PM Subject: Re: [ovirt-users] self hosted setup with answer file On Mon, Feb 16, 2015 at 1:18 AM, Yedidyah Bar David <didi(a)redhat.com&gt; wrote: > ----- Original Message ----- > > From: "Greg Swift" <gregswift(a)gmail.com&gt; > > To: users(a)ovirt.org > > Sent: Monday, February 16, 2015 7:37:53 AM > > Subject: Re: [ovirt-users] self hosted setup with answer file > > > > I'm starting to get more concerned. It seems almost every time i run it > off > > the same answer file I get a different setup. > > > > On Sun, Feb 15, 2015 at 11:45 AM, Greg Swift < gregswift(a)gmail.com > > wrote: > > > > > > > > So i've been playing with standing up a self hosted instance of oVirt. > I'm > > using an answer file that i generated first time I tried running it (so > that > > I can provision w/ ansible) > > > > I've run into 2 problems. > > > > 1: i have to specify the jboss home manually from the cli (didn't see an > > answer equivalent), even though the ovirt jboss package is installed. > > Are you referring to 'hosted-engine --deploy' on the host, to > 'engine-setup' > inside the VM, both? You need a different answer file for each, and the one > generated for 'hosted-engine --deploy' is currently not enough for > automatic > unattended installation. For that, see [1]. > > [1] https://bugzilla.redhat.com/show_bug.cgi?id=1101297 > So I was installing the allinone plugin and then running engine-setup. > > > > > 2: Generally speaking it seems to be ignoring several of the answer file > > config entries. My answers tell it to 'confirm settings' and it runs > > non-interactively. Plus all the local storage configs seem to be ignored. > > > > I'm running ovirt-engine-3.5.1.1-1.el7.centos.noarch > > > > Any suggestions? > > In addition to the above, if you suspect a real bug, logs and answer files > will help. > So the problem is that at this point I've killed the system that wasn't working and I think I found a solution. I wish i had thought to save off the logs, but at least in theory it seemed to be re-produceable. That being said the following changes to my original answer file (OP attachment) and I've got a working environment -OVESETUP_CONFIG/applicationMode=str:both +OVESETUP_CONFIG/applicationMode=str:virt -OVESETUP_CONFIG/storageIsLocal=bool:True +OVESETUP_CONFIG/storageIsLocal=bool:False -OVESETUP_SYSTEM/nfsConfigEnabled=bool:True +OVESETUP_SYSTEM/nfsConfigEnabled=bool:False -OVESETUP_CONFIG/isoDomainMountPoint=str:/var/lib/exports/iso -OVESETUP_CONFIG/isoDomainName=str:ISO_DOMAIN -OVESETUP_CONFIG/isoDomainACL=str:{{ ansible_fqdn }}(rw) +OVESETUP_CONFIG/isoDomainMountPoint=none:None +OVESETUP_CONFIG/isoDomainName=none:None +OVESETUP_CONFIG/isoDomainACL=none:None Some of that is just weird though... storageIsLocal=False? its completely local storage.

But hey.. its working consistently now.