On Tue, Jan 21, 2020 at 8:37 AM <webmattr(a)hotmail.com> wrote:
Hello,
I can't seem to install the self-hosted engine onto local storage. It
gives me glustefs, iscsi, fc, and nfs as the available options. I'm using
this in a home-lab scenario, and don't have budget/etc. for building out a
dedicated NAS for it, or setting up multiple nodes. I like the look of
oVirt, and wanted to try it with a couple disposable vm's (plex, and a
docker instance I break often). My current best-thought for how to make it
work is to setup NFS on the server, and then point the self-hosted engine
at the (local) NFS share. Is there a better way to do this that I might be
overlooking?*
*Factoring that I don't have the funds to build out a proper storage
environment, yet.
(and if anyone asks, I did search for a solution to this, but didn't find
anything super helpful. Mostly I found 5+ year old articles on a similar
but different scenario).
You can run an hyperconverged infrastructure where you actually have only
one node connecting via gluster to itself.
Now it is fully supported and no need to use the workaround of having NFS
running locally
Main upstream doc here:
https://www.ovirt.org/documentation/self-hosted/chap-Deploying_Self-Hoste...
Main downstream doc with more details here:
https://access.redhat.com/documentation/en-us/red_hat_hyperconverged_infr...
deployment workflow here:
https://access.redhat.com/documentation/en-us/red_hat_hyperconverged_infr...
and after initial host configuration steps you will arrive at the step of
cockpit configuration where you choose the "Hyperconverged" option (at the
right) and then button "Run Gluster Wizard For Single Node":
https://access.redhat.com/documentation/en-us/red_hat_hyperconverged_infr...
HIH,
Gianluca