Jump to: navigation, search

Difference between revisions of "Mellanox-Neutron-Kilo-Ubuntu-InfiniBand"

(Controller Node)
 
(16 intermediate revisions by 3 users not shown)
Line 2: Line 2:
  
 
==Mellanox Neutron ML2 Driver==
 
==Mellanox Neutron ML2 Driver==
 +
  
 
Mellanox ML2 Mechanism Driver implements the ML2 Plugin Mechanism Driver API.  
 
Mellanox ML2 Mechanism Driver implements the ML2 Plugin Mechanism Driver API.  
  
This driver supports Mellanox embedded switch functionality as part of the VPI (Ethernet/InfiniBand) HCA. Mellanox ML2 Mechanism Driver provides functional parity with Mellanox Neutron plugin.  
+
This driver supports Mellanox embedded switch functionality as part of the InfiniBand HCA. Mellanox ML2 Mechanism Driver provides functional parity with Mellanox Neutron plugin.  
  
Mellanox ML2 Mechanism Driver supports DIRECT (pci passthrough) vnic type. For vnic type configuration API details, please refer to configuration reference guide (click here). Hardware vNICs mapped to the guest VMs allow higher performance and advanced features such as RDMA (remote direct memory access).  
+
Mellanox ML2 Mechanism Driver supports DIRECT (pci passthrough) vnic type. For vnic type configuration API details, please refer to configuration reference guide.
 +
Hardware vNICs mapped to the guest VMs allow higher performance and advanced features such as RDMA (remote direct memory access).  
  
The driver supports VLAN network type to facilitate virtual networks either on Ethernet or InfiniBand fabrics.  
+
The driver supports VLAN network type to facilitate virtual networks either on InfiniBand fabrics.  
 
* Mellanox OpenStack Neutron Agent (L2 Agent) runs on each compute node.  
 
* Mellanox OpenStack Neutron Agent (L2 Agent) runs on each compute node.  
 
* Agent should apply VIF connectivity based on mapping between a VIF (VM vNIC) and Embedded Switch port.  
 
* Agent should apply VIF connectivity based on mapping between a VIF (VM vNIC) and Embedded Switch port.  
  
 
===Prerequisites===
 
===Prerequisites===
* A running OpenStack environment installed with the ML2 plugin on top of Linux Bridge.
+
* A running OpenStack environment installed with the ML2 plugin on top of OpenVswitch or Linux Bridge.
 
* All nodes equipped with Mellanox ConnectX®-3 Network Adapter (http://www.mellanox.com/page/products_dyn?product_family=119)  
 
* All nodes equipped with Mellanox ConnectX®-3 Network Adapter (http://www.mellanox.com/page/products_dyn?product_family=119)  
 
* Mellanox OFED 2.4 or greater installed on all nodes. Please refer to Mellanox website for the latest OFED: http://www.mellanox.com/page/products_dyn?product_family=26&mtag=linux_sw_drivers
 
* Mellanox OFED 2.4 or greater installed on all nodes. Please refer to Mellanox website for the latest OFED: http://www.mellanox.com/page/products_dyn?product_family=26&mtag=linux_sw_drivers
Line 20: Line 22:
 
* The software package iproute2 - (http://www.linuxfoundation.org/collaborate/workgroups/networking/iproute2 ) installed on all Compute nodes
 
* The software package iproute2 - (http://www.linuxfoundation.org/collaborate/workgroups/networking/iproute2 ) installed on all Compute nodes
 
* Add repository  
 
* Add repository  
<pre> sudo add-apt-repository http://www.mellanox.com/repository/solutions/openstack/kilo/ubuntu/ </pre>
+
sudo add-apt-repository http://www.mellanox.com/repository/solutions/openstack/kilo/ubuntu
  
 
=InfiniBand Network=
 
=InfiniBand Network=
 
The Mellanox Neutron Plugin use InfiniBand Partitions (PKeys) to separate Networks.  
 
The Mellanox Neutron Plugin use InfiniBand Partitions (PKeys) to separate Networks.  
  
==SM Configuration==
+
==SM Node==
  
 
=== OpenSM Provisioning with SDN Mechanism Driver ===
 
=== OpenSM Provisioning with SDN Mechanism Driver ===
Line 37: Line 39:
 
To configure OpenSM:  
 
To configure OpenSM:  
  
1. Make sure that all the PKeys are predefined in the partitions.conf file ('''/etc/opensm/partitions.conf''').
+
1. Make sure that all the PKeys are predefined in the '''/etc/opensm/partitions.conf'''.
 
 
2. Add/Change the following in the partitions.conf file:
 
 
  management=0xffff,ipoib, sl=0, defmember=full: ALL, ALL_SWITCHES=full,SELF=full;
 
  management=0xffff,ipoib, sl=0, defmember=full: ALL, ALL_SWITCHES=full,SELF=full;
  
3. For every network you want to configure in Neutron, you have to configure the PKey associated with the VLAN of this network (defined in Neutron): vlan1=0x1, ipoib, sl=0, defmember=full : ALL;
+
2. For every network you want to configure in Neutron, you have to configure the PKey associated with the VLAN of this network (defined in Neutron): vlan1=0x1, ipoib, sl=0, defmember=full : ALL;
  
 
Below is an example of the configuration of the partitions.conf file in case you have 10 VLANs defined in the configuration of the file:
 
Below is an example of the configuration of the partitions.conf file in case you have 10 VLANs defined in the configuration of the file:
 
'''/etc/neutron/plugins/mlnx/mlnx_conf.ini'''. ( network_vlan_ranges = physnet1:1:10)  
 
'''/etc/neutron/plugins/mlnx/mlnx_conf.ini'''. ( network_vlan_ranges = physnet1:1:10)  
<pre>
+
management=0x7fff,ipoib, sl=0, defmember=full : ALL, ALL_SWITCHES=full,SELF=full;
management=0x7fff,ipoib, sl=0, defmember=full : ALL, ALL_SWITCHES=full,SELF=full;
+
vlan1=0x1, ipoib, sl=0, defmember=full: ALL_CAS;
vlan1=0x1, ipoib, sl=0, defmember=full: ALL_CAS;
+
vlan2=0x2, ipoib, sl=0, defmember=full: ALL_CAS;
vlan2=0x2, ipoib, sl=0, defmember=full: ALL_CAS;
+
vlan3=0x3, ipoib, sl=0, defmember=full: ALL_CAS;
vlan3=0x3, ipoib, sl=0, defmember=full: ALL_CAS;
+
vlan4=0x4, ipoib, sl=0, defmember=full: ALL_CAS;
vlan4=0x4, ipoib, sl=0, defmember=full: ALL_CAS;
+
vlan5=0x5, ipoib, sl=0, defmember=full: ALL_CAS;
vlan5=0x5, ipoib, sl=0, defmember=full: ALL_CAS;
+
vlan6=0x6, ipoib, sl=0, defmember=full: ALL_CAS;
vlan6=0x6, ipoib, sl=0, defmember=full: ALL_CAS;
+
vlan7=0x7, ipoib, sl=0, defmember=full: ALL_CAS;
vlan7=0x7, ipoib, sl=0, defmember=full: ALL_CAS;
+
vlan8=0x8, ipoib, sl=0, defmember=full: ALL_CAS;
vlan8=0x8, ipoib, sl=0, defmember=full: ALL_CAS;
+
vlan9=0x9, ipoib, sl=0, defmember=full: ALL_CAS;
vlan9=0x9, ipoib, sl=0, defmember=full: ALL_CAS;
+
vlan10=0xa, ipoib, sl=0, defmember=full: ALL_CAS;
vlan10=0xa, ipoib, sl=0, defmember=full: ALL_CAS;
 
</pre>
 
  
 
4. Modify the following line in the file '''/etc/opensm/opensm.conf''' from FALSE to TRUE:  
 
4. Modify the following line in the file '''/etc/opensm/opensm.conf''' from FALSE to TRUE:  
<pre>
+
allow_both_pkeys TRUE
allow_both_pkeys TRUE
 
</pre>
 
  
 
5. Restart the OpenSM:  
 
5. Restart the OpenSM:  
<pre>
+
# service opensmd restart
#service opensmd restart
 
</pre>
 
  
 
==Controller Node==
 
==Controller Node==
  
To configure the Controller node:  
+
To configure the Controller node:
1.Configure yum with Mellanox OpenStack Kilo repository
+
1.Configure Mellanox OpenStack Kilo repository
<pre>
+
# sudo echo "deb [arch=amd64] http://www.mellanox.com/repository/solutions/openstack/kilo/ubuntu/14.04 openstack-mellanox main" >> /etc/apt/sources.list.d/mellanox-openstack-repository.list
#sudo add-apt-repository http://www.mellanox.com/repository/solutions/openstack/kilo/ubuntu
+
# sudo apt-get update
</pre>
+
 
 +
2. Install Mellanox deb packages:
 +
# apt-get install -y neutron-plugin-mlnx python-networking-mlnx
 +
 
 +
3. Make sure ML2 is the current Neutron plugin by checking the core_plugin parameter in '''/etc/neutron/neutron.conf''':
 +
core_plugin = neutron.plugins.ml2.plugin.Ml2Plugin
 +
 
 +
4. Make sure '''/etc/neutron/plugin.ini''' is pointing to '''/etc/neutron/plugins/ml2/ml2_conf.ini''' (symbolic link)
 +
 
 +
5. Modify /etc/neutron/plugins/ml2/ml2_conf.ini by adding the following:
 +
[ml2]
 +
type_drivers = vlan,flat
 +
tenant_network_types = vlan
 +
mechanism_drivers = mlnx, openvswitch
 +
# or mechanism_drivers = mlnx, [linuxbridge|openvswitch]
 +
[ml2_type_vlan]
 +
network_vlan_ranges = physnet1:1:10
 +
[eswitch]
 +
# (StrOpt) Type of Network Interface to allocate for VM:
 +
# mlnx_direct or hostdev according to libvirt terminology
 +
vnic_type = hostdev
 +
 
 +
6. Start (or restart) the Neutron server:
 +
# service neutron-server restart
 +
 
 +
== Network Node==
 +
To configure the Network node:
  
2. Install Mellanox RPMs:  
+
Prerequisites:
<pre>
+
python-libxml2 package is installed
modify gpgcheck=0 in /etc/yum.conf
+
# sudo apt-get install python-libxml2
#apt-get install -y openstack-neutron-mellanox eswitchd python-networking-mlnx mlnx-dnsmasq
 
</pre>
 
  
3. Edit the following file: /usr/lib/systemd/system/neutron-mlnx-agent.service
+
1. Make sure that eIPoIB module is up and configured in '''/etc/infiniband/openib.conf''': For more information, please refer to eIPoIB configuration in Mellanox OFED User Manual.  
<pre>
+
E_IPOIB_LOAD=yes
Change the  /etc/neutron/plugins/mlnx/mlnx.ini to /etc/neutron/plugins/mlnx/mlnx_conf.ini
 
</pre>
 
  
4. Run:
 
<pre>
 
#service enable neutron-mlnx-agent.service
 
</pre>
 
  
5. Run:  
+
2. Restart openibd:  
<pre>
+
  # service openibd restart
  #service daemon-reload
 
</pre>
 
  
6. Make sure ML2 is the current Neutron plugin by checking the core_plugin parameter in /etc/neutron/neutron.conf:
 
<pre>
 
core_plugin = neutron.plugins.ml2.plugin.Ml2Plugin
 
</pre>
 
  
7. Make sure '''/etc/neutron/plugin.ini''' is pointing to '''/etc/neutron/plugins/ml2/ml2_conf.ini''' (symbolic link)
+
3. Modify the network bridge configuration according to the use of OpenVswitch or LinuxBridge
 +
* 3.1 OpenVswitch '''/etc/neutron/plugins/openvswitch/ovs_neutron_plugin.ini'''  
 +
[ovs]
 +
bridge_mappings = physnet1:br-<eIPoIB interface>
 +
* 3.2 LinuxBridge file '''/etc/neutron/plugins/linuxbridge/linuxbridge_conf.ini''':
 +
[linux_bridge]
 +
physical_interface_mappings = physnet1:<eIPoIB interface>
  
8. Modify '''/etc/neutron/plugins/ml2/ml2_conf.ini''' by adding the following:
 
<pre>
 
[ml2]
 
type_drivers = vlan,flat
 
tenant_network_types = vlan
 
mechanism_drivers = mlnx, openvswitch
 
# or mechanism_drivers = mlnx, linuxbridge
 
[ml2_type_vlan]
 
network_vlan_ranges = physnet1:1:10
 
[eswitch]
 
# (StrOpt) Type of Network Interface to allocate for VM:
 
# mlnx_direct or hostdev according to libvirt terminology
 
vnic_type = hostdev
 
</pre>
 
9. Start (or restart) the Neutron server:
 
#service neutron-server restart
 
  
==Compute Nodes==
+
NOTE: In order to obtain the eIPoIB interface name, run the ethtool tool command (see below) and check that driver name is eth_ipoib:
 +
# ethtool -i <eIPoIB_interface>
 +
driver: eth_ipoib
 +
.....
  
To configure the Compute Node:
 
  
1.Configure yum with Mellanox OpenStack Kilo repository
+
4. Restart network bridge and neutron-dhcp-agent:
<pre>
+
# service neutron-dhcp-agent restart
#yum-config-manager --add-repo http://www.mellanox.com/repository/solutions/openstack/kilo/ubuntu
+
* 4.1 OpenVswitch
</pre>
+
# service neutron-linuxbridge-agent restart
 +
* 4.2 LinuxBridge
 +
# service neutron-openvswitch-agent restart
  
2. Install Mellanox RPMs:
+
NOTE: For DHCP support, the Network node should use the Mellanox Dnsmasq driver as the DHCP driver.  
<pre>
 
modify gpgcheck=0 in /etc/yum.conf
 
#yum install -y openstack-neutron-mellanox eswitchd python-networking-mlnx mlnx-dnsmasq
 
</pre>
 
  
3. Edit the following file: '''/usr/lib/systemd/system/neutron-mlnx-agent.service'''
+
===DHCP Server (Usually part of the Network node)===
<pre>
+
1. Install mlnx_dnsmasq:
Change the /etc/neutron/plugins/mlnx/mlnx.ini to /etc/neutron/plugins/mlnx/mlnx_conf.ini
+
  # sudo apt-get -y --force-yes install mlnx-dnsmasq
</pre>
 
  
4. Run:  
+
2. Modify '''/etc/neutron/dhcp_agent.ini''' as follows and according to OVS or Linuxbridge:  
<pre>
+
dhcp_driver = mlnx_dhcp.MlnxDnsmasq
  #systemctl enable neutron-mlnx-agent.service
+
dhcp_broadcast_reply = True
</pre>
+
  # or interface_driver = neutron.agent.linux.interface.BridgeInterfaceDriver
 +
interface_driver = neutron.agent.linux.interface.OVSInterfaceDriver
  
5. Run:
 
<pre>
 
#systemctl daemon-reload
 
</pre>
 
  
6. Apply MLNX patch for Kilo:
+
3. Start DHCP server:  
<pre>
+
  # service neutron-dhcp-agent restart
  #pushd /usr/lib/python2.7/site-packages/nova/virt/libvirt/
 
#wget http://www.mellanox.com/repository/solutions/openstack/kilo/ubuntu/patch/mlnx_kilo.patch
 
#patch < mlnx_kilo.patch
 
#popd
 
</pre>
 
  
7. Create the file '''/etc/modprobe.d/mlx4_ib.conf''' and add the following: 
+
==Compute Nodes==
<pre>
 
options mlx4_ib sm_guid_assign=0
 
</pre>
 
  
8. Restart Nova:  
+
To configure the Compute Node:  
<pre>
 
#systemctl restart openstack-nova-compute
 
</pre>
 
  
9. Restart the driver:
+
1.Configure Mellanox OpenStack Kilo repository
<pre>
+
# sudo add-apt-repository http://www.mellanox.com/repository/solutions/openstack/kilo/ubuntu
# /bin/systemctl restart opensm.service
+
# apt-get update
# /etc/init.d/openibd restart
 
</pre>
 
  
10. In the file '''/etc/neutron/plugins/mlnx/mlnx_conf.ini''', the parameters tenant_network_type , and network_vlan_ranges should be configured as the controllers:
+
2. Install Mellanox deb packages:
<pre>
+
# apt-get install -y neutron-plugin-mlnx-agent neutron-plugin-mlnx eswitchd python-networking-mlnx
physical_interface_mappings = physnet1:<ib_interface>(for example physnet1:ib0)
 
</pre>
 
  
11. Modify the file '''/etc/eswitchd/eswitchd.conf''' as follows:
+
3. Apply MLNX patch for Kilo:
<pre>
+
# pushd /usr/lib/python2.7/site-packages/nova/virt/libvirt/
  fabrics = physnet1:<ib_interface> (for example physnet1:ib0)
+
# wget http://www.mellanox.com/repository/solutions/openstack/kilo/patch/mlnx_kilo.patch
</pre>
+
# patch < mlnx_kilo.patch
 +
  # popd
  
12. Start eSwitchd:  
+
4. Add a missing line to network filters:
<pre>
+
  # echo "ebrctl: CommandFilter, ebrctl, root" >> /etc/nova/rootwrap.d/compute.filters
#systemctl restart eswitchd
 
</pre>
 
  
13. Start the Neutron agent:  
+
5. Verify correct name of Mellanox agent service:
<pre>
+
  # sed -i s/neutron-plugin-mlnx-agent/neutron-mlnx-agent/g /etc/init/neutron-plugin-mlnx-agent.conf
#systemctl restart neutron-mlnx-agent
 
</pre>
 
  
== Network Node==
+
6. In the file '''/etc/neutron/plugins/mlnx/mlnx_conf.ini''', the parameters tenant_network_type , and network_vlan_ranges should be configured as the controllers:
To configure the Network node:
+
physical_interface_mappings = physnet1:<ib_interface>(for example physnet1:ib0)
  
1. Make sure that eIPoIB module is up and configured in '''/etc/infiniband/openib.conf''': For more information, please refer to eIPoIB configuration in Mellanox OFED User Manual.
+
7. Modify the file '''/etc/eswitchd/eswitchd.conf''' as follows:
E_IPOIB_LOAD=yes
+
fabrics = physnet1:<ib_interface> (for example physnet1:ib0)
2. Restart openibd:
 
#service openibd restart
 
3. Modify the file '''/etc/neutron/plugins/linuxbridge/linuxbridge_conf.ini''' as follows:
 
<pre>
 
[linux_bridge]
 
physical_interface_mappings = physnet1:<eIPoIB interface>
 
</pre>
 
if you are using openvswitch than modify /etc/neutron/plugins/openvswitch/ovs_neutron_plugin.ini
 
<pre>
 
[ovs]
 
bridge_mappings = physnet1:br-<eIPoIB interface>
 
</pre>
 
  
NOTE: In order to obtain the eIPoIB interface name, run the ethtool tool command (see below) and check that driver name is eth_ipoib:  
+
8. Start eSwitchd:  
  #ethtool -i <eIPoIB_interface>
+
  # initctl start eswitchd
driver: eth_ipoib
 
version: 1.0.0
 
firmware-version: 1
 
bus-info: ib0
 
supports-statistics: yes
 
supports-test: no
 
supports-eeprom-access: no
 
supports-register-dump: no
 
supports-priv-flags: no
 
  
4. Restart neutron-linuxbridge-agent and neutron-dhcp-agent:  
+
9. Start the Neutron agent:  
  #systemctl restart neutron-linuxbridge-agent.service
+
  # initctl start neutron-plugin-mlnx-agent
#systemctl restart neutron-openvswitch-agent.service
 
  
NOTE: For DHCP support, the Network node should use the Mellanox Dnsmasq driver as the DHCP driver.
+
10. Restart Nova compute:  
===DHCP Server (Usually part of the Network node)===
+
  # service nova-compute restart
1. Modify /etc/neutron/dhcp_agent.ini as follows and according to OVS or Linuxbridge:  
 
<pre>
 
dhcp_driver = mlnx_dhcp.MlnxDnsmasq
 
  dhcp_broadcast_reply = True
 
#or interface_driver = neutron.agent.linux.interface.BridgeInterfaceDriver
 
interface_driver = neutron.agent.linux.interface.OVSInterfaceDriver
 
</pre>
 
2. Start DHCP server:
 
#systemctl restart neutron-dhcp-agent.service
 
  
 
==Known issues and Troubleshooting==
 
==Known issues and Troubleshooting==
 
For known issues and troubleshooting options refer to [https://community.mellanox.com/docs/DOC-1127 Mellanox OpenStack Troubleshooting]
 
For known issues and troubleshooting options refer to [https://community.mellanox.com/docs/DOC-1127 Mellanox OpenStack Troubleshooting]

Latest revision as of 16:02, 22 March 2016

Overview

Mellanox Neutron ML2 Driver

Mellanox ML2 Mechanism Driver implements the ML2 Plugin Mechanism Driver API.

This driver supports Mellanox embedded switch functionality as part of the InfiniBand HCA. Mellanox ML2 Mechanism Driver provides functional parity with Mellanox Neutron plugin.

Mellanox ML2 Mechanism Driver supports DIRECT (pci passthrough) vnic type. For vnic type configuration API details, please refer to configuration reference guide. Hardware vNICs mapped to the guest VMs allow higher performance and advanced features such as RDMA (remote direct memory access).

The driver supports VLAN network type to facilitate virtual networks either on InfiniBand fabrics.

  • Mellanox OpenStack Neutron Agent (L2 Agent) runs on each compute node.
  • Agent should apply VIF connectivity based on mapping between a VIF (VM vNIC) and Embedded Switch port.

Prerequisites

sudo add-apt-repository http://www.mellanox.com/repository/solutions/openstack/kilo/ubuntu

InfiniBand Network

The Mellanox Neutron Plugin use InfiniBand Partitions (PKeys) to separate Networks.

SM Node

OpenSM Provisioning with SDN Mechanism Driver

SDN Mechanism Driver allows OpenSM dynamically assign PKs in the IB network.

More details about applying SDN Mechanism Driver with NEO can be found here

Manual OpenSM Configuration

To configure OpenSM:

1. Make sure that all the PKeys are predefined in the /etc/opensm/partitions.conf.

management=0xffff,ipoib, sl=0, defmember=full: ALL, ALL_SWITCHES=full,SELF=full;

2. For every network you want to configure in Neutron, you have to configure the PKey associated with the VLAN of this network (defined in Neutron): vlan1=0x1, ipoib, sl=0, defmember=full : ALL;

Below is an example of the configuration of the partitions.conf file in case you have 10 VLANs defined in the configuration of the file: /etc/neutron/plugins/mlnx/mlnx_conf.ini. ( network_vlan_ranges = physnet1:1:10)

management=0x7fff,ipoib, sl=0, defmember=full : ALL, ALL_SWITCHES=full,SELF=full;
vlan1=0x1, ipoib, sl=0, defmember=full: ALL_CAS;
vlan2=0x2, ipoib, sl=0, defmember=full: ALL_CAS;
vlan3=0x3, ipoib, sl=0, defmember=full: ALL_CAS;
vlan4=0x4, ipoib, sl=0, defmember=full: ALL_CAS;
vlan5=0x5, ipoib, sl=0, defmember=full: ALL_CAS;
vlan6=0x6, ipoib, sl=0, defmember=full: ALL_CAS;
vlan7=0x7, ipoib, sl=0, defmember=full: ALL_CAS;
vlan8=0x8, ipoib, sl=0, defmember=full: ALL_CAS;
vlan9=0x9, ipoib, sl=0, defmember=full: ALL_CAS;
vlan10=0xa, ipoib, sl=0, defmember=full: ALL_CAS;

4. Modify the following line in the file /etc/opensm/opensm.conf from FALSE to TRUE:

allow_both_pkeys TRUE

5. Restart the OpenSM:

# service opensmd restart

Controller Node

To configure the Controller node: 1.Configure Mellanox OpenStack Kilo repository

# sudo echo "deb [arch=amd64] http://www.mellanox.com/repository/solutions/openstack/kilo/ubuntu/14.04 openstack-mellanox main" >> /etc/apt/sources.list.d/mellanox-openstack-repository.list
# sudo apt-get update

2. Install Mellanox deb packages:

# apt-get install -y neutron-plugin-mlnx python-networking-mlnx

3. Make sure ML2 is the current Neutron plugin by checking the core_plugin parameter in /etc/neutron/neutron.conf:

core_plugin = neutron.plugins.ml2.plugin.Ml2Plugin

4. Make sure /etc/neutron/plugin.ini is pointing to /etc/neutron/plugins/ml2/ml2_conf.ini (symbolic link)

5. Modify /etc/neutron/plugins/ml2/ml2_conf.ini by adding the following:

[ml2]
type_drivers = vlan,flat
tenant_network_types = vlan
mechanism_drivers = mlnx, openvswitch
# or mechanism_drivers = mlnx, [linuxbridge|openvswitch]
[ml2_type_vlan]
network_vlan_ranges = physnet1:1:10
[eswitch]
# (StrOpt) Type of Network Interface to allocate for VM:
# mlnx_direct or hostdev according to libvirt terminology
vnic_type = hostdev

6. Start (or restart) the Neutron server:

# service neutron-server restart

Network Node

To configure the Network node:

Prerequisites:

python-libxml2 package is installed
# sudo apt-get install python-libxml2

1. Make sure that eIPoIB module is up and configured in /etc/infiniband/openib.conf: For more information, please refer to eIPoIB configuration in Mellanox OFED User Manual.

E_IPOIB_LOAD=yes


2. Restart openibd:

# service openibd restart


3. Modify the network bridge configuration according to the use of OpenVswitch or LinuxBridge

  • 3.1 OpenVswitch /etc/neutron/plugins/openvswitch/ovs_neutron_plugin.ini
[ovs]
bridge_mappings = physnet1:br-<eIPoIB interface>
  • 3.2 LinuxBridge file /etc/neutron/plugins/linuxbridge/linuxbridge_conf.ini:
[linux_bridge] 
physical_interface_mappings = physnet1:<eIPoIB interface>


NOTE: In order to obtain the eIPoIB interface name, run the ethtool tool command (see below) and check that driver name is eth_ipoib:

# ethtool -i <eIPoIB_interface> 
driver: eth_ipoib
.....


4. Restart network bridge and neutron-dhcp-agent:

# service neutron-dhcp-agent restart
  • 4.1 OpenVswitch
# service neutron-linuxbridge-agent restart
  • 4.2 LinuxBridge
# service neutron-openvswitch-agent restart

NOTE: For DHCP support, the Network node should use the Mellanox Dnsmasq driver as the DHCP driver.

DHCP Server (Usually part of the Network node)

1. Install mlnx_dnsmasq:

 # sudo apt-get -y --force-yes install mlnx-dnsmasq

2. Modify /etc/neutron/dhcp_agent.ini as follows and according to OVS or Linuxbridge:

dhcp_driver = mlnx_dhcp.MlnxDnsmasq
dhcp_broadcast_reply = True
# or interface_driver = neutron.agent.linux.interface.BridgeInterfaceDriver
interface_driver = neutron.agent.linux.interface.OVSInterfaceDriver


3. Start DHCP server:

# service neutron-dhcp-agent restart

Compute Nodes

To configure the Compute Node:

1.Configure Mellanox OpenStack Kilo repository

  1. sudo add-apt-repository http://www.mellanox.com/repository/solutions/openstack/kilo/ubuntu
  2. apt-get update

2. Install Mellanox deb packages:

# apt-get install -y neutron-plugin-mlnx-agent neutron-plugin-mlnx eswitchd python-networking-mlnx

3. Apply MLNX patch for Kilo:

# pushd /usr/lib/python2.7/site-packages/nova/virt/libvirt/ 
# wget http://www.mellanox.com/repository/solutions/openstack/kilo/patch/mlnx_kilo.patch
# patch < mlnx_kilo.patch
# popd

4. Add a missing line to network filters:

 # echo "ebrctl: CommandFilter, ebrctl, root" >> /etc/nova/rootwrap.d/compute.filters

5. Verify correct name of Mellanox agent service:

 # sed -i s/neutron-plugin-mlnx-agent/neutron-mlnx-agent/g /etc/init/neutron-plugin-mlnx-agent.conf

6. In the file /etc/neutron/plugins/mlnx/mlnx_conf.ini, the parameters tenant_network_type , and network_vlan_ranges should be configured as the controllers:

physical_interface_mappings = physnet1:<ib_interface>(for example physnet1:ib0)

7. Modify the file /etc/eswitchd/eswitchd.conf as follows:

fabrics = physnet1:<ib_interface> (for example physnet1:ib0)

8. Start eSwitchd:

# initctl start eswitchd

9. Start the Neutron agent:

# initctl start neutron-plugin-mlnx-agent

10. Restart Nova compute:

# service nova-compute restart

Known issues and Troubleshooting

For known issues and troubleshooting options refer to Mellanox OpenStack Troubleshooting