Hi Gianluca, The work flow you described seems correct to me. You may want to run rpmconf after step 5 just to be sure you're not missing some configuration changes. -- Sandro Il 22/02/2013 06:45, Gianluca Cecchi ha scritto:

Suppose that for your Fedora 18 oVirt 3.2 all-in-one config, where you have both the engine (ovirt-engine service) and the node (vdsmd service) there are a bunch of updates available such as: vdsm kernel ...

is it correct to go through the workflow below?

1) stop running VMs

2) upgrade engine 2a) no related updates # engine-upgrade

Checking for updates... (This may take several minutes)...[ DONE ] No updates available

2b) some related updates (eg. Supposing x.y below is the same as the one currrently installed

Checking for updates... (This may take several minutes)...[ DONE ] 11 Updates available: * ovirt-engine-x.y.x-4.fc18.noarch * ovirt-engine-backend-x.y.z-4.fc18.noarch ..

Error: New ovirt-engine-setup rpm available via yum. Please execute `yum update ovirt-engine-setup`, then re-execute 'engine-upgrade'. )

# yum update ovirt-engine-setup # engine-upgrade

3) stop engine # systemctl stop ovirt-engine.service

4) stop vdsmd # systemctl stop vdsmd.service

5) update remaining system packages if something already done in step 2

# yum update

6) reboot and check all is ok

Gianluca _______________________________________________ Users mailing list Users@ovirt.org http://lists.ovirt.org/mailman/listinfo/users