Note: This bug is displayed in read-only format because the product is no longer active in Red Hat Bugzilla.

Bug 1168242

Summary:	packstack no longer tap the provider ovs-bridge to a NIC.
Product:	Red Hat OpenStack	Reporter:	Nir Magnezi <nmagnezi>
Component:	openstack-packstack	Assignee:	Lukas Bezdicka <lbezdick>
Status:	CLOSED ERRATA	QA Contact:	Nir Magnezi <nmagnezi>
Severity:	high	Docs Contact:
Priority:	high
Version:	6.0 (Juno)	CC:	aortega, derekh, gchamoul, mmagr, nyechiel, oblaut, sclewis, ssainkar, tfreger, yeylon
Target Milestone:	ga	Keywords:	AutomationBlocker, Regression
Target Release:	6.0 (Juno)
Hardware:	Unspecified
OS:	Linux
Whiteboard:
Fixed In Version:	openstack-packstack-2014.2-0.13.dev1395.gaabe0a2.el7ost	Doc Type:	Bug Fix
Doc Text:	Packstack did not tap the provider ovs-bridge to a NIC. Packstack set only first tenant network type specified in CONFIG_NEUTRON_ML2_TENANT_NETWORK_TYPES. This meant that for example CONFIG_NEUTRON_ML2_TENANT_NETWORK_TYPES=vxlan,vlan packstack did set up only the vxlan. This is now fixed and packstack does configuration of both tenant network types.	Story Points:	---
Clone Of:	1163697	Environment:
Last Closed:	2015-02-09 14:53:38 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:	1163697
Bug Blocks:

Description Nir Magnezi 2014-11-26 13:34:49 UTC

Reproduced with RHEL-OSP6-Beta: 
openstack-packstack-puppet-2014.2-0.5.dev1316.g733aa73.el7ost.noarch
openstack-packstack-2014.2-0.5.dev1316.g733aa73.el7ost.noarch

+++ This bug was initially created as a clone of Bug #1163697 +++

Description of problem:
=======================
This seems like a regression with packstack.
I configured the answer file with provider ovs bridge, yet the bridge (named br-nodes) was tapped to the relevant NIC in the networker node but not in the compute nodes.
The motivation for this test is to allow a provider IPv6 vlan network.
In addition Used another NIC as a tunnel interface, to allow vxlan/gre tenant networks as well (for non-provider networks).

snipped from answer file:
CONFIG_NEUTRON_L3_EXT_BRIDGE=provider
CONFIG_NEUTRON_L2_PLUGIN=ml2
CONFIG_NEUTRON_ML2_TYPE_DRIVERS=vxlan,vlan
CONFIG_NEUTRON_ML2_TENANT_NETWORK_TYPES=vxlan,vlan
CONFIG_NEUTRON_ML2_MECHANISM_DRIVERS=openvswitch
CONFIG_NEUTRON_ML2_VLAN_RANGES=ipv6_vlan_range:60:66
CONFIG_NEUTRON_ML2_TUNNEL_ID_RANGES=1:1000
CONFIG_NEUTRON_ML2_VNI_RANGES=10:100
CONFIG_NEUTRON_L2_AGENT=openvswitch
CONFIG_NEUTRON_OVS_TENANT_NETWORK_TYPE=vxlan,vlan
CONFIG_NEUTRON_OVS_VLAN_RANGES=ipv6_vlan_range:60:66
CONFIG_NEUTRON_OVS_BRIDGE_MAPPINGS=ipv6_vlan_range:br-nodes
CONFIG_NEUTRON_OVS_BRIDGE_IFACES=br-nodes:enp5s0f1
CONFIG_NEUTRON_OVS_TUNNEL_RANGES=1:1000
CONFIG_NEUTRON_OVS_TUNNEL_IF=enp4s0f1

Version-Release number of selected component (if applicable):
=============================================================
RDO-Juno.
openstack-packstack-2014.2-0.5.dev1316.g733aa73.el7.centos.noarch
openstack-packstack-puppet-2014.2-0.5.dev1316.g733aa73.el7.centos.noarch

How reproducible:
=================
1/1

Steps to Reproduce:
===================
1. Used the attached answer file with packstack.
2. post installation, check your compute nodes for the provider bridge ports:
   # ovs-vsctl show

Actual results:
===============
    Bridge br-nodes
        Port phy-br-nodes
            Interface phy-br-nodes
                type: patch
                options: {peer=int-br-nodes}
        Port br-nodes
            Interface br-nodes
                type: internal

Expected results:
=================
    Bridge br-nodes
        Port "enp5s0f1"
            Interface "enp5s0f1"
        Port phy-br-nodes
            Interface phy-br-nodes
                type: patch
                options: {peer=int-br-nodes}
        Port br-nodes
            Interface br-nodes
                type: internal
        Port "enp5s0f1\nbr-nodes"
            Interface "enp5s0f1\nbr-nodes"

Additional Info:
================
Workround: ovs-vsctl add-port br-nodes enp5s0f1

Comment 6 Nir Magnezi 2014-12-14 12:16:06 UTC

Reopening.
Got the exact same result with: openstack-packstack-2014.2-0.12.dev1370.ge46bee1.el7ost.noarch
See logs attached

Comment 15 Lukas Bezdicka 2015-01-06 15:21:05 UTC

The original issue with packstack not tapping devices to the bridges when CONFIG_NEUTRON_ML2_TYPE_DRIVERS=vxlan,gre,vlan was set is fixed and works, the second issue with not tapping when vlan is missing from CONFIG_NEUTRON_ML2_TYPE_DRIVERS is not a bug.

Comment 17 Lukas Bezdicka 2015-01-12 10:15:42 UTC

I'm sorry my last comment was wrong, again:
CONFIG_NEUTRON_ML2_TYPE_DRIVERS=vxlan,gre,vlan is comma separated list of network type driver entrypoints to be loaded from the neutron.ml2.type_drivers namespace. The original bug was about CONFIG_NEUTRON_ML2_TENANT_NETWORK_TYPES=vxlan,gre,vlan not setting up vlan.

The CONFIG_NEUTRON_ML2_TENANT_NETWORK_TYPES is comma separated ordered list of network_types to allocate as tenant networks. And if there was more than one option in list packstack previously didn't set up anything than the first value.

Comment 18 Nir Magnezi 2015-01-12 13:22:37 UTC

Verified NVR: openstack-packstack-2014.2-0.13.dev1395.gaabe0a2.el7ost.noarch

Following to comment #15, Tested OK.
Following to that Bug I filed Bug #1181131

Comment 20 errata-xmlrpc 2015-02-09 14:53:38 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-2015-0155.html