On 08/02/2012 09:47 AM, Neil wrote:
Hi guys,
Thank you very much for all your input and assistance.
On Tue, Jul 31, 2012 at 5:48 PM, Ofer Schreiber <oschreib@redhat.com> wrote:
it seems that you're using a very old 3.1 build, which we never released during the beta process (afair).
this build is a x86_64 build, which confuses the upgrade script (since new rpms are noarch).
I can't guarantee it will work fine (although I can't think of any reason it won't), but using the --force-setup... parameter is a good option.
Actually, we already fixed this bug upstream, but I'm not sure it will get into 3.1.
I'll try the --force-setup today if I get a chance, in a worst case scenario what could go wrong? When I run the upgrade will it create a DB backup that can be manually restored if I have to remove all ovirt-engine packages and start from scratch? Anyone have some steps on how to go about restoring from the backup that the "engine-upgrade --force-setup" creates?
Currently I'm running a two node/hypervisor setup with a SAN as my data storage, but I can't migrate VM's between the nodes due to what appears to be a bug, which is the reason for wanting to update the ovirt-engine. As previously mentioned my engine is installed directly on one of the nodes/hypervisors.
though not directly related, that's not helping...
Would it perhaps be a better idea to install a new ovirt-engine onto a different server and then migrate to the new engine rather than updating the current one, or am I just being completely paranoid?
We are running mission critical VM's on the install, so I can't risk something going wrong and the VM's stop running or access to them is lost.
upgrade wasn't tested thourougly in these versions. installing a new environment and moving VMs to it would be safer, but it will require you to have more nodes, and stop/export/import/launch your VMs.
Thank you very much.
Regards.
Neil Wilson _______________________________________________ Users mailing list Users@ovirt.org http://lists.ovirt.org/mailman/listinfo/users