Neutron SDN Manual Implementation Manual

Introduction to Installation Architecture

The purpose of this paper is to build an environment of SDN, such as Neutron (Openvswitch + GRE), to learn how it works, to simulate the core principles, such as: the same tenant custom network instance interoperability, manually for instance distribution floating IP and other phase Content. <ignore_js_op>

Virtual network

You need to create a new 3 virtual network Net0, Net1, and Net2, which are configured in virtual box as follows.

1. Net0:
2. Network Name:virtualbox host-only Ethernet adapter#2
3. Purpose:administrator/management Network
4. IP block:10.20.0.0/24
5. Dhcp:disable
6. Linux Device:eth0
8. Net1:
9. Network Name:virtualbox host-only Ethernet adapter#3
10. Purpose:public Network
11. Dhcp:disable
12. IP block:172.16.0.0/24
13. Linux device:eth1
15. Net2:
16. Network Name:virtualbox host-only Ethernet adapter#4
17. Purpose:storage/private Network
18. Dhcp:disable
19. IP block:192.168.4.0/24
20. Linux Device:eth2

Copy Code

Virtual machines

You need to create a new 2 virtual machine VM1 and VM2, which are configured as follows.

1. VM1:
2. Name:network1
3. Vcpu:1
4. memory:1g
5. disk:30g
6. Network:net1,net2,net3
8. VM2:
9. Name:compute1
10. Vcpu:1
11. memory:1g
12. disk:30g
13. Networks:net1,net2,net3

Copy Code

Linux Interface Settings

1. Network1
2. eth0:10.20.0.201 (Management Network)
3. eht1:172.16.0.201 (Public/external Network)
4. eht2:192.168.4.201 (Private Network,gre tunning)
6. Compute1
7. ETH0:10.20.0.202 (Management Network)
8. EHT1: (Disabled)
9. eht2:192.168.4.202 (Private Network,gre tunning)

Copy Code

Simulating the installation of network nodes (NETWORK1)

Analog Network node-related implementation, such as L3, dhcp-agent implementation, in order to simulate the multi-node network situation, here also simulates a compute node, the simulation M2 Openvswitch implementation, the above run Instance1.

Network Interface Configuration

1. Vi/etc/sysconfig/network-scripts/ifcfg-eth0
2. Device=eth0
3. Type=ethernet
4. Onboot=yes
5. Nm_controlled=yes
6. Bootproto=static
7. ipaddr=10.20.0.201
8. netmask=255.255.255.0
10. Vi/etc/sysconfig/network-scripts/ifcfg-eth1
11. Device=eth1
12. Type=ethernet
13. Onboot=yes
14. Nm_controlled=yes
15. Bootproto=static
16. ipaddr=172.16.0.201
17. netmask=255.255.255.0
19. Vi/etc/sysconfig/network-scripts/ifcfg-eth2
20. Device=eth2
21. Type=ethernet
22. Onboot=yes
23. Nm_controlled=yes
24. Bootproto=static
25. ipaddr=192.168.4.201
26. netmask=255.255.255.0

Copy Code

Restart Network Service

1. Service Network restart

Copy Code

Install the package that needs to be used

1. Yum install libvirt openvswitch python-virtinst Xauth tigervnc-y

Copy Code

Remove the default Libvirt network for easy and clear analysis of network conditions

1. Virsh Net-destroy Default
2. Virsh Net-autostart--disable Default
3. Virsh net-undefine Default

Copy Code

Set Allow Ipforwarding

1. Vi/etc/sysctl.conf
2. Net.ipv4.ip_forward=1
3. Net.ipv4.conf.all.rp_filter=0
4. Net.ipv4.conf.default.rp_filter=0

Copy Code

Immediate effect

1. Sysctl-p

Copy Code

Start Openvswitch

1. Service Openvswitch Start
2. Chkconfig Openvswitch on

Copy Code

Create a Linux Bridge

1. Brctl ADDBR Qbr01
2. IP link Set Qbr01 up

Copy Code

Create a instance and connect to Qbr01 Bridge, the network Interface section is configured as follows

1. <interface type= ' bridge ' >
2. <source bridge= ' Qbr01 '/>
3. <target dev= ' tap01 '/>
4. <model type= ' Virtio '/>
5. <driver name= ' Qemu '/>
6. <address type= ' PCI ' domain= ' 0x0000 ' bus= ' 0x00 ' slot= ' 0x03 ' function= ' 0x0 '/>
7. </interface>

Copy Code

You can refer to the attachment./gre/instance1.xml Create

1. CP ~/gre//var/tmp/
2. Cd/var/tmp/gre
3. MV Cirros-0.3.0-x86_64-disk.img instance1.img
4. Virsh Define Instance1.xml
5. Virsh start Instance1
6. Virsh Vncdisplay Instance1
7. vncviewer:0

Copy Code

After you start the console, login to add the IP address 192.168.1.11

1. IP addr Add 192.168.1.11/24 dev eth0
2. Route add default GW 192.168.1.1

Copy Code

Create an internal bridge br-int, simulate OpenStack integrated bridge

1. Ovs-vsctl ADD-BR Br-int
2. Ovs-vsctl add-port br-int gre0--set interface gre0 Type=gre options:remote_ip=192.168.4.202

Copy Code

Create a Veth peer, connect the Linux bridge ' Qbr01 ' and Openvswich Bridge ' Br-ini '

1. IP link Add qvo01 type Veth peer name Qvb01
2. Brctl addif Qbr01 QVB01
3. Ovs-vsctl Add-port Br-int qvo01
4. Ovs-vsctl Set Port qvo01 tag=100
5. IP link Set qvb01 up
6. IP link Set qvo01 up

Copy Code

Check out the Br-int on NETWORK1 now.

1. Ovs-vsctl Show

Copy Code

Simulating installation compute nodes (COMPUTE1)


Network Interface Configuration

1. Vi/etc/sysconfig/network-scripts/ifcfg-eth0
2. Device=eth0
3. Type=ethernet
4. Onboot=yes
5. Nm_controlled=yes
6. Bootproto=static
7. ipaddr=10.20.0.202
8. netmask=255.255.255.0
10. Vi/etc/sysconfig/network-scripts/ifcfg-eth1
11. Device=eth1
12. Type=ethernet
13. Onboot=yes
14. Nm_controlled=yes
15. Bootproto=static
16. ipaddr=172.16.0.202
17. netmask=255.255.255.0
19. Vi/etc/sysconfig/network-scripts/ifcfg-eth2
20. Device=eth2
21. Type=ethernet
22. Onboot=yes
23. Nm_controlled=yes
24. Bootproto=static
25. ipaddr=192.168.4.202
26. netmask=255.255.255.0

Copy Code

Restart Network Service

1. Service Network restart

Copy Code

Install the package that needs to be used

1. Yum install libvirt openvswitch python-virtinst Xauth tigervnc

Copy Code

Remove the Libvirt default network

1. Virsh Net-destroy Default
2. Virsh Net-autostart--disable Default
3. Virsh net-undefine Default

Copy Code

Set Allow Ipforwarding

1. Vi/etc/sysctl.conf
2. Net.ipv4.ip_forward=1
3. Net.ipv4.conf.all.rp_filter=0
4. Net.ipv4.conf.default.rp_filter=0

Copy Code

Immediate effect

1. Sysctl-p

Copy Code

Start Openvswitch

1. Service Openvswitch Start
2. Chkconfig Openvswitch on

Copy Code

Create a Linux Bridge

1. Brctl ADDBR Qbr02
2. IP link Set qbr02 up

Copy Code

Create a VM and connect to the Qbr02

On the GRE directory to the COMPUTE1 node, you can refer to the attachment./gre/instance2.xml Create

1. CP ~/gre//var/tmp/
2. Cd/var/tmp/gre
3. MV Cirros-0.3.0-x86_64-disk.img instance2.img
4. Virsh Define Instance2.xml
5. Virsh Start Instance2
6. Virsh Vncdesplay Instance2
7. vncviewer:0

Copy Code

After you start the console, login to add IP to know 192.168.1.12

1. IP addr Add 192.168.1.12/24 dev eth0
2. Route add default GW 192.168.1.1

Copy Code

Create an internal bridge br-int, simulate OpenStack integrated bridge

1. Ovs-vsctl ADD-BR Br-int
2. Ovs-vsctl add-port br-int gre0--set interface gre0 Type=gre options:remote_ip=192.168.4.201

Copy Code

Create a Veth peer, connect the Linux bridge ' Qbr02 ' and Openvswich Bridge ' Br-ini '

1. IP link Add qvo02 type Veth peer name qvb02
2. Brctl addif Qbr02 QVB02
3. Ovs-vsctl Add-port Br-int qvo02
4. Ovs-vsctl Set Port qvo02 tag=100
5. IP link Set qvb02 up
6. IP link Set qvo02 up

Copy Code

Check out the Br-int on NETWORK1 now.

1. Ovs-vsctl Show

Copy Code

Check if the Instance1 can be connected to the INSTANCE2 console

1. Ping 192.168.1.11

Copy Code

Access to private network of tenants through networking Namespace

Add a namespace,dhcp01 to isolate the tenant network.

1. IP netns Add dhcp01

Copy Code

192.168.1.0/24 for a private network, creating a DHCP service in the namespace Dhcp01

1. Ovs-vsctl add-port br-int TAPDHCP01--set interface Tapdhcp01 type=internal
2. Ovs-vsctl Set Port Tapdhcp01 tag=100
4. IP link set tapdhcp01 netns dhcp01
5. IP netns exec dhcp01 IP addr add 192.168.1.2/24 dev tapdhcp01
6. IP netns exec dhcp01 IP link set tapdhcp01 up

Copy Code

Check network connectivity, Access Instance1 and Instance2 in namespace

1. IP netns exec dhcp01 ping 192.168.1.12
2. IP netns exec dhcp01 ping 192.168.1.11

Copy Code

L3 router through Network Namespace and iptables

Ovs-vsctl ADD-BR Br-ex

Reconfigure Eth1 and Br-ex

1. Vi/etc/sysconfig/network-scripts/ifcfg-eth1
3. Device=eth1
4. Onboot=yes
5. Bootproto=none
6. Promisc=yes
7. mtu=1546
9. Vi/etc/sysconfig/network-scripts/ifcfg-br-ex
11. Device=br-ex
12. Type=bridge
13. Onboot=yes
14. Bootproto=none
15. ipaddr0=172.16.0.201
16. Prefix0=24

Copy Code

Restart the Network Service

1. Ovs-vsctl add-port br-ex eth1 && Service Network restart

Copy Code

Check the network for connectivity after configuration

1. Ping 172.16.0.201

Copy Code

Add a Namespace,router01 for routing and floating IP assignment

1. IP netns Add Router01

Copy Code

Add an interface to Br-int as a gateway to the private network 192.168.1.0/24

1. Ovs-vsctl add-port br-int Qr01--set interface Qr01 type=internal
2. Ovs-vsctl Set Port Qr01 tag=100
4. IP link set Qr01 netns Router01
5. IP netns exec router01 IP addr add 192.168.1.1/24 dev Qr01
6. IP netns exec router01 IP link set qr01 up
7. IP netns exec router01 IP link set lo up

Copy Code

Add an interface in Br-ex for private network 192.168.1.0/24 set next hop address

1. Ovs-vsctl add-port Br-ex QG01--set interface qg01 type=internal
2. IP link set qg01 netns Router01
3. IP netns exec router01 IP addr add 172.16.0.100/24 dev qg01
4. IP netns exec router01 IP link set qg01 up
5. IP netns exec router01 IP link set lo up

Copy Code

Analog assignment floating IP access Instance1

assigning floating ip,172.16.0.101 to Instance1 192.168.1.11

1. IP netns exec router01 IP addr add 172.16.0.101/32 dev qg01
3. IP netns exec router01 iptables-t nat-a output-d 172.16.0.101/32-j DNAT--to-destination 192.168.1.11
4. IP netns exec router01 iptables-t nat-a prerouting-d 172.16.0.101/32-j DNAT--to-destination 192.168.1.11
5. IP netns exec router01 iptables-t nat-a postrouting-s 192.168.1.11/32-j SNAT--to-source 172.16.0.101
6. IP netns exec router01 iptables-t nat-a postrouting-s 192.168.1.0/24-j SNAT--to-source 172.16.0.100

Copy Code

Test floating IP

1. Ping 172.16.0.101

Copy Code

If you need to clear NAT chain

1. Iptables-t nat-f

Neutron SDN Manual Implementation Manual