Bug 1325025

Summary:	It should be specificed clearly that undercloud deployment should be done in a Virtual machine alone.
Product:	Red Hat OpenStack	Reporter:	Ruchika K <rkharwar>
Component:	documentation	Assignee:	RHOS Documentation Team <rhos-docs>
Status:	CLOSED WONTFIX	QA Contact:	RHOS Documentation Team <rhos-docs>
Severity:	medium	Docs Contact:
Priority:	unspecified
Version:	7.0 (Kilo)	CC:	byount, dtantsur, kbasil, mburns, srevivo
Target Milestone:	---	Keywords:	Documentation
Target Release:	10.0 (Newton)
Hardware:	All
OS:	Linux
Whiteboard:
Fixed In Version:		Doc Type:	Bug Fix
Doc Text:		Story Points:	---
Clone Of:		Environment:
Last Closed:	2016-11-25 14:12:02 UTC	Type:	Bug
Regression:	---	Mount Type:	---
Documentation:	---	CRM:
Verified Versions:		Category:	---
oVirt Team:	---	RHEL 7.3 requirements from Atomic Host:
Cloudforms Team:	---	Target Upstream Version:

Description Ruchika K 2016-04-07 22:07:44 UTC

Description of problem:

Undercloud desplyoment should be done in a Vm alone. We should tighten up the options on this unless there is a really good reason to leave it open for baremetal.

This is the justification to do an undercloud install on a VM alone

* "undercloud uninstall" does not exist so there is no planned and supported way to uninstall an undercloud.
* if customers break it in a non recoverable manner and want to start clean state there is NO way to do it on baremetal.

On baremetal.
* Uninstalling triple o and re-installing it does not clean up the state of the undercloud
* Reinstalling undercloud does not fix the state 
* The only option is to reimage the server which is NOT practical in a customer environment since customers do not want to reconfigure their networking etc and it costs them some time.

In a Virtual Machine
* The customer merely needs to revert to the snapshot before the undercloud was installed.


How reproducible:
NA

Steps to Reproduce:
1.
2.
3.

Actual results:


Expected results:


Additional info:

Comment 2 Mike Burns 2016-09-30 08:32:24 UTC

I don't disagree with documenting the undercloud as an option (see bug 1337772).  I do believe that we should enumerate some of the pros and cons of the 2 approaches and allow customers to make an informed choice.

Comment 3 Dmitry Tantsur 2016-11-25 14:12:02 UTC

Hello. While it's *probably* possible to even deploy undercloud in a VM for bare metal overcloud, this not something we ever test, to say nothing about supporting. TripleO is expected to be used on bare metal, and so it Ironic. We are aware of limitations of several virtual solutions. Also, to be honest, undercloud is resource-greedy enough to deserve a separate BM machine.

As https://bugzilla.redhat.com/show_bug.cgi?id=1337772 already covers documenting it as an option, I'm closing this request.