Note: This bug is displayed in read-only format because the product is no longer active in Red Hat Bugzilla.

Bug 1364478

Summary:	[RFE] Support adding ceph storage networks to the compute nodes
Product:	Red Hat OpenStack	Reporter:	John Fulton <johfulto>
Component:	openstack-tripleo-heat-templates	Assignee:	Giulio Fidente <gfidente>
Status:	CLOSED ERRATA	QA Contact:	Yogev Rabl <yrabl>
Severity:	unspecified	Docs Contact:
Priority:	unspecified
Version:	10.0 (Newton)	CC:	dbecker, gfidente, jschluet, mburns, morazi, rhel-osp-director-maint
Target Milestone:	rc	Keywords:	FutureFeature, TestOnly, Triaged
Target Release:	10.0 (Newton)
Hardware:	Unspecified
OS:	Unspecified
Whiteboard:
Fixed In Version:	openstack-tripleo-heat-templates-5.0.0-0.20160929150845.4cdc4fc.el7ost	Doc Type:	Enhancement
Doc Text:	Feature: Allow usage of any isolated network on any role. Reason: Some scenarios, like the HCI deployment where 'ceph-osd' is colocated with 'nova-compute', assume that nodes have access to multiple isolated networks, like it happens for the controller nodes. Result: Custom NIC templates can now configure any of the isolated network on any role.	Story Points:	---
Clone Of:		Environment:
Last Closed:	2016-12-14 15:49:21 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:
Embargoed:
Bug Depends On:
Bug Blocks:	1291946

Description John Fulton 2016-08-05 12:56:32 UTC

In previous releases, the compute role contains the same networks as
the ceph storage role with the exception of the storage management
network [1]. In Liberty this could be achieved by modifying
advanced-networking.yaml [2]. In Newton all ports are now wired into
all nodes but not enabled [3]. The RFE requests that a user may easily
enable them in Director for OSP10 should the user choose to deploy
OSDs on the same nodes as computes (hyperconverged as requested in RH BZ
1291946). In order to implement this the overcloud image will need to
be built with the puppet to enable this port [4] and the compute
configuration must happen after the controllers. 

[1] 
Controller:
- Provisioning
- Internal API
- Storage
- Storage Management
- External

Compute:
- Provisioning
- Internal API
- Storage
- Tenant

Ceph Storage:
- Provisioning
- Storage
- Storage Management

[2] Add the following to templates/advanced-networking.yaml
OS::TripleO::Compute::Ports::StoragePort: /home/stack/templates/network/ports/storage.yaml

[3] https://review.openstack.org/#/c/338113

[4] https://review.openstack.org/#/c/340286

Comment 2 John Fulton 2016-09-20 12:58:19 UTC

A new upstream submission will be coming in for review with lines 3 and 4 of the following uncommented: 

 https://review.openstack.org/#/c/338113/10/environments/hyperconverged-ceph.yaml

The user should then be able to: 

openstack overcloud deploy ... -e ~/templates/environments/hyperconverged-ceph.yaml ...

Submission will then be linked here.

Comment 3 Giulio Fidente 2016-09-23 13:57:45 UTC

The HCI environment file (environments/hyperconverged-ceph.yaml) will default to use of NETISO on the compute nodes.

Comment 4 John Fulton 2016-10-03 16:16:21 UTC

FYI: environments/hyperconverged-ceph.yaml depends on python-heatclient 1.5 or newer. As per https://review.rdoproject.org/r/#/c/2857 said heatclient is merged in RDO (however there's a slight upstream divergence here which should get fixed too https://review.rdoproject.org/r/#/c/2857).

Comment 7 Yogev Rabl 2016-11-24 08:20:16 UTC

Verified on OSP 10 openstack-tripleo-heat-templates-5.1.0-3.el7ost.noarch

Comment 9 errata-xmlrpc 2016-12-14 15:49:21 UTC

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.

https://rhn.redhat.com/errata/RHEA-2016-2948.html