[Users] proliant errors trying to install ovirt-node image

Mon Mar 5 12:49:52 UTC 2012

> -----Original Message-----
> From: users-bounces at ovirt.org [mailto:users-bounces at ovirt.org] On
> Behalf Of ext Joe Pruett
> Sent: Sunday, March 04, 2012 2:11 AM
> To: users at ovirt.org
> Subject: [Users] proliant errors trying to install ovirt-node image
> 
> i've already found the info on the list about the "global name 'disk'
> not defined" error and i've tried both an iso from the jenkins url
> listed, as well as creating my own git area using the 2.2.3 tagged code
> and making the one line change to fix that python error.  but neither
> of
> those images work either.   i get a generic "ovirt node hypervisor
> installation failed" message.  i've poked around the log viewers and at
> the end i see it complaining that the gpt partition label has no bios
> bootable partition, but it isn't clear if that is really the issue.
I ran into the same problems as you when trying to install on a
HP ProLiant BL460c G1 blade.

The Red Hat Bugzilla report #798030 "TUI Installation fails with
 Grub error" seems to address exactly this kind of problem.

The grub2-setup complaining has got its reason: In file script/storage.py
the disk partitioning is done using a GPT, but no partition with 
flag bios_grub is created.

I modified the source (mis-)using the first partition (EFI partition)
in case of a non-EFI boot as a bios_grub flagged partition. This
partition is much bigger than necessary (grub2.info recommends one
or two MB), thus leaving some room for optimization for the experts when
fixing the problem.

Please find the patch that allowed successful installation in the
attachment.

> i'm
> wondering if there might be more cciss driver issues.
> 
> i haven't dug into the source to figure out the root pw so i could more
> easily look at the log files or export them.  can someone enlighten me
> with that more quickly than i can find it?
As I am stuck to reading log files online or exporting them via screen shots,
I am also interested in learning about better ways of exporting. 

> 
> what is the best procedure to follow to try and get to the bottom of
> issues like this?  might i be better off installing a regular os and
> then installing the vdsm package?
Installing Fedora 16 on the same kind of blade worked for me without any
problems. Therefore the "regular os + vdsm rpms" approach could help
you to sidestep the failing ovirt-node installation, but it will probably
not help you in analyzing and fixing ovirt-node related installation issues.

Best regards,

Hermann Huy

St.-Martin-Str. 76
81541 München
Tel: +49 89 5159 35744
Mob: +49 170 5630316
Email: hermann.huy at nsn.com
http://www.nokiasiemensnetworks.com/global/

Nokia Siemens Networks GmbH & Co. KG
Sitz der Gesellschaft: München / Registered office: Munich
Registergericht: München / Commercial registry: Munich, HRA 88537
WEEE-Reg.-Nr.: DE 52984304

Persönlich haftende Gesellschafterin / General Partner: Nokia Siemens Networks Management GmbH
Geschäftsleitung / Board of Directors: Hermann Rodler, Olaf Horsthemke
Vorsitzender des Aufsichtsrats / Chairman of supervisory board: Herbert Merz
Sitz der Gesellschaft: München / Registered office: Munich
Registergericht: München / Commercial registry: Munich, HRB 163416

-------------- next part --------------
A non-text attachment was scrubbed...
Name: 0001-added-bios_grub-partition.patch
Type: application/octet-stream
Size: 2878 bytes
Desc: 0001-added-bios_grub-partition.patch
URL: <http://lists.ovirt.org/pipermail/users/attachments/20120305/0623789a/attachment-0001.obj>