2

I have a oVirt 4.2 datacenter. I like to upgrade to 4.3. What steps are necessary and where do I get documentation. The oVirt documentation is sparse on this matter to say the least.

6

To answer my own question:

There is good documentation available from Redhat for RHEV 4.3.

Don't use any Redhat repos. Stay with the release notes of oVirt and use their repos only, which are CentOS based.

If anything in this answer is unclear to you: RTFM until clearing up happens.

I recommend waiting for 4.3.4 to arrive, since 4.3.3 is still a little bit rough on the edges.

This guide is for oVirt Node NG based installations. If you rolled out your own nodes this guide may be working as well, but I don't know. Have backups of your engine. REALLY!

I did:

1. Check of all nodes are using firewalld. You have to switch all nodes to use firewalld, because Iptables is deprecated and removed in 4.3 as it was said in documentation. You have to re-install every node for that. So every node has to go through maintenance.

2. Upgrade the engine. First to the 4.2.8 then to 4.3.x

My upgrade flow for this is:

Engine Update START

Minor Upgrades first

enable global maintenance mode

login to engine
engine-upgrade-check
yum update "ovirt-*-setup*"
engine-setup [The questions come with sane defaults from your current installation]
yum update
when success: disable global maintenance mode

Major Upgrade

enable global maintenance mode

login to engine
yum install http://resources.ovirt.org/pub/yum-repo/ovirt-release[releasenumber].rpm
engine-upgrade-check
yum update "ovirt-*-setup*"
engine-setup [Same questions like above, but also for new features]
remove the old ovirt release from /etc/yum.repos.d
yum update
when success: disable global maintenance mode

Engine Upgrade END

3. Upgrade the nodes.

First to 4.2.8 then as specified in ovirt release info:

  # yum install https://resources.ovirt.org/pub/ovirt-4.3/rpm/el7/noarch/ovirt-node-ng-image-update-4.3.3-1.el7.noarch.rpm
  # reboot

4. Upgrade the Datacenter Compatibility to 4.3 This includes rebooting all VMs at one point. You can read about it in the documentation.

Then you're done.

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.