[ovirt-users] Re: Hyperconverged setup ovirt 4.3.x using ansible?

I cannot really answer your question, but I have a similar one and some observations. I also wanted to split the three node HC setup into the Gluster part and the HC magic engine injection. I have been using Glusters in combination with a non-hosted or local engine with success and I thought it better to go step-by-step in case something goes wrong and you have to fix it: Under-the-hood is only good until something goes wrong, right? But there is no evident script for the HC single or three node setups and the ansible jobs underneath those wizards are at least not advertised and documented from what I could see. However once I did go through with the three node wizard and encountered failures during the hosted-engine setup, I noticed that it would recognise a matching Gluster setup and forward immediately to the hosted-engine part. And on for just how that Gluster has to look to make the HC setup like it, you may have to study all the Ansible playbooks and scripts in /usr/src. Because just any Gluster can be used using the "existing storage" variant or hosted-engine-setup script, but will lack the full depth of integration and evidently preclude the use of Gluster nodes as compute nodes, as they are already part of another Gluster and Gluster precludes multiple memberships. Don't know if this helps...