[ovirt-users] Rebuilding my infra..
carl langlois
crl.langlois at gmail.com
Mon Jan 8 20:32:59 UTC 2018
I should have say replica 3+arbiter when adding the 3rd host.
Yes i know that the documentation specified 3 hosts. But if i understand
this is to make sure the system is always reliable and that make sense. But
my questions is mainly is it possible to start with 2 nodes and had a 3rd
one later..when i had the budget. If i am not mistaking Gluster can work
with 2 nodes.
On Mon, Jan 8, 2018 at 3:15 PM, Vinícius Ferrão <ferrao at if.ufrj.br> wrote:
> If I’m not wrong GlusterFS in oVirt requires 3 hosts.
>
> Here’s the RHHI guide, it’s pretty much the same for oVirt:
> https://access.redhat.com/documentation/en-us/red_hat_
> hyperconverged_infrastructure/1.1/html/deploying_red_hat_
> hyperconverged_infrastructure/
>
> > On 8 Jan 2018, at 18:10, carl langlois <crl.langlois at gmail.com> wrote:
> >
> > Hi all
> >
> > After screwing my infra with the update to 4.2 (probably a bad
> manipulation), i am planning a rebuild of the entire infra. First i want to
> replace my NFS storage with a glusterfs storage. All documentation tell me
> that i need 3 hosts.. but for the moment i only have 2 but planning to had
> more later.
> >
> > So does it make sense to start with 2 hosts and use glusterfs as the
> storage domain(lets says with a replicate of two with all its limitations).
> > If it make sense,
> > 1- what is the best way to do it.
> > 2- how hard will it be to had the 3rd host when available and make it
> replica 2+arbiter.
> >
> > Also in a setup where i have 3 hosts (replica 2+arbiter) does all the 3
> hosts can run users vm?
> >
> > Thanks for your inputs.
> >
> > Carl
> >
> > _______________________________________________
> > Users mailing list
> > Users at ovirt.org
> > http://lists.ovirt.org/mailman/listinfo/users
>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.ovirt.org/pipermail/users/attachments/20180108/30a0a93c/attachment.html>
More information about the Users
mailing list