[Deepsea-users] How to "undo" a complete installation using deepsea

Jose Betancourt jose.betancourt at suse.com
Tue Jan 24 08:50:02 MST 2017

Hi Eric,

Thank you for your quick turnaround.  I did the deepsea installation following section 4.2 of the SES4 installation guide via zypper.   If you have the rpm version of the packages that you mentioned below, let me know their location and I'll give them a try.

Thanks again,

José Betancourt
Linux Architecture - IHV Alliances & Embedded Systems
jose.betancourt at suse.com

-----Original Message-----
From: deepsea-users-bounces at lists.suse.com [mailto:deepsea-users-bounces at lists.suse.com] On Behalf Of Eric Jackson
Sent: Tuesday, January 24, 2017 5:36 AM
To: Discussions about the DeepSea management framework for Ceph <deepsea-users at lists.suse.com>
Subject: Re: [Deepsea-users] How to "undo" a complete installation using deepsea

Hi José,

On Tuesday, January 24, 2017 05:24:23 AM Jose Betancourt wrote:
> Hello Everyone,
> I have a small setup with 6 physical devices.  I've been using deepsea 
> and probably ran salt-run state.orch ceph.stage.configure and 
> ceph.stage.deploy too many times and it's now complaining that I have 
> too few monitors and storage nodes.

I'm curious, but to your question...
> What I would like to do is to be able to roll back to a point where I can
> start over.   With ceph-deploy, I have the option of running ceph-deploy
> purge, purgedata and forgetkeys and I can pretty much start again 
> (this is a lab environment).
> Is there an equivalent salt-run invocation to basically "reset" the 
> salt-run steps so that I can start at stage 1 again?

I have added ceph.purge in https://github.com/SUSE/DeepSea/pull/78.  I completed this Saturday and it effectively removes the Ceph cluster and resets DeepSea.  I am waiting for feedback from another developer, but you are welcome to try it.  

The commands would be

salt-run disengage.safety
salt-run state.orch ceph.purge

and optionally

salt 'admin*' purge.proposals

That third step can be included in the default ceph.purge.  That's one of the questions I was hoping for feedback.  (The default removes the cluster and most of the pillar configuration and allows you to start at Stage 2.  If your Stage 1 will remain the same, then removing the proposals doesn't really add

Are you cloning from github or working from an rpm?  If the latter, I will do another release of master as soon as this branch is merged.


> Best,
> José Betancourt
> Linux Architecture - IHV Alliances & Embedded Systems 
> jose.betancourt at suse.com<mailto:jose.betancourt at suse.com>
> +1.908.672.2719

More information about the Deepsea-users mailing list