Description of problem:
For Ceph cluster deployment OSP deployer (Staypuft) MUST be able to PXE boot a node that Ceph will use and install RHELOSP RHEL7 operating system with all Ceph server library dependencies (for block, object, NAS (in priority order)),
setup networks (bonds, vLANs) per node type (storage node, controller node - see https://bugzilla.redhat.com/show_bug.cgi?id=1119874),
ability to specify host group separately for MON and OSD nodes (for network setup).
Ability to modify networks configuration for nodes that Ceph is using.
Version-Release number of selected component (if applicable):
Steps to Reproduce:
nobt sure if RHELOSP5 RHEL7 should include all Ceph libraries.
Then ICE will handle configuration and setup portion of Ceph deployment
ICE will install all the relevant Ceph components. There is no need for the RHEL-OSP Installer to do this.
[The RHEL-OSP Installer should only handle installing the Ceph client libraries on any Glancer/Cinder/Nova nodes as necessary.]
Also, my understanding (which could be wrong) is that Staypuft just refers to the UI. It is really Foreman that needs to do the heavy lifting for this BZ.
*** Bug 1122667 has been marked as a duplicate of this bug. ***
New Ceph host group will be added that will do *only* base RHEL 7 provisioning
To elaborate on the initial effort:
2 new host groups will be added
Ceph Storage Node (OSD)
Generic RHEL 7
These will be exposed in the Assign Hosts screen and will allow network configuration, etc.
The backend puppet classes used will be core RHEL 7 only.
No Ceph Storage configuration will be done. That is the domain of the ICE installer.
PR submitted: https://github.com/theforeman/staypuft/pull/272
Brad, could you double check the doc-text edit? Just want to ensure I understood your intent. thanks, Summer
do we have a kickstart file specific to generic RHEL7 and to Ceph Node (OSD)?
(In reply to arkady kanevsky from comment #12)
> do we have a kickstart file specific to generic RHEL7 and to Ceph Node (OSD)?
There is only 1 kickstart. There will be (in the next release) a condition around the partitioning for ceph nodes, otherwise, it's the same kickstart.
verified in version ruby193-rubygem-staypuft-0.3.9-1.el6ost.noarch
The Staypuft has installed a clean RHEL 7.
From what I understand this is the requirement at this stage of the installation.
Since the problem described in this bug report should be
resolved in a recent advisory, it has been closed with a
resolution of ERRATA.
For information on the advisory, and where to find the updated
files, follow the link below.
If the solution does not work for you, open a new bug report.