Hello,

I am running oVirt 4.3.6 and glusterd service and peers on all HCI nodes are connected and working properly after updating.  I'm not sure if I understand what your question or issue is.  It definitely should be safe to update your oVirt HCI to 4.3.6.  GlusterFS is able to update from 5.6 to 6.5 one host at a time, the other hosts are still able to connect after one has been updated to Gluster 6x

On Tue, Oct 1, 2019 at 9:52 AM Satheesaran Sundaramoorthi <sasundar@redhat.com> wrote:
On Tue, Oct 1, 2019 at 6:00 PM Jayme <jaymef@gmail.com> wrote:
In an oVirt HCI gluster configuration you should be able to take down at least one host at a time.  The procedure I use to upgrade my oVirt HCI cluster goes something like this:

1. Upgrade the oVirt engine setup packages.  Upgrade the engine.  Yum update and reboot.
2. Place the first host in maintenance mode and upgrade the host.  Reboot.  Wait until the host is active and gluster bricks are fully healed. 
3. Place the next host in maintenance and perform the same steps until all hosts are upgraded. 

From: "Akshita Jain" <akshita@councilinfosec.org>
Subject: Unable to Upgrade
Date: 1 October 2019 at 11:12:58 CEST

After upgrading oVirt 4.3.4 to 4.3.6, the gluster is also upgrading from 5.6 to 6.5. But as soon as it upgrades gluster peer status shows disconnected.
What is the correct method to upgrade oVirt with gluster HCI environment?

Hello Akshita,

Jayme's steps are perfect enough for the update.
If you see post updating to oVirt 4.3.6, your peers are in disconnected state,
could you confirm that glusterd service is up and running on all HCI nodes ( # systemctl status glusterd ).

Meanwhile, I can also try for the quick update from gluster-5.6 to 6.5 and update the results here.

-- Satheesaran S ( sas )