[rdo-list] Packstack refactor and future ideas

[Ivan Chavero]

----- Original Message -----
> From: "Alan Pevec" <apevec at redhat.com>
> To: "James Slagle" <jslagle at redhat.com>
> Cc: "rdo-list" <rdo-list at redhat.com>, "Javier Pena" <javier.pena at redhat.com>
> Sent: Thursday, June 9, 2016 9:50:44 AM
> Subject: Re: [rdo-list] Packstack refactor and future ideas
> 
> > the current discussion is about making AIO better or making multinode
> > easier via packstack.
> 
> Current discussion is about scoping Packstack right, so it can both
> fulfill its mission to be an easy introduction to OpenStack for the
> new community members and at the same time have maintainable codebase.
> "An easy introduction to OpenStack" means providing real working cloud
> on a normal 8-16 GB laptop with the ability to run at least one real
> instance, not cirros in emulated VM-in-VM.
> For production use, as you say, doing multinode right is not simple
> and that's why there's TripleO.
> Multinode that has been historically included in Packstack is
> incomplete for real production usage and making it better is out of
> scope hence my proposal to remove it.


Taking in account comments by Javier and Daniel, i think we can effectively
eliminate the multinode functionality of Packstack but not loosing its ability
to install different components in a separate way.

Extending what Daniel said we could do stuff like:

packstack --controller-node
packstack --compute-node
packstack --network-node

and if you want you can use:

packstack --os-install-mariadb=y , etc...

Cheers,
Ivan