| Re: IPv6 and OVZ part deux [message #41437 is a reply to message #41436] |
Fri, 21 January 2011 04:24   |
lars.bailey
Messages: 38 Registered: April 2010
|
Member |
|
|
@ Jean-Marc
This is what I now use on a test Node,for IPv6/OVZ/Ethernet bridging.
The Node server;
* /etc/sysconfig/network
NETWORKING=yes
HOSTNAME=stooge
GATEWAYDEV=virtbr0
NETWORKING_IPV6=yes
IPV6INIT=yes
IPV6FORWARDING=yes
IPV6_AUTOCONF=no
IPV6_AUTOTUNNEL=no
This is the source-route bridge network configuration.
* ifcfg-virtbr0
DEVICE=virtbr0
TYPE=Bridge
ONBOOT=yes
STP=off
DELAY=0
BOOTPROTO=static
IPADDR=192.168.1.72
NETMASK=255.255.255.0
IPV6ADDR=fd22:a075:afd0:e096::101/64
IPV6_DEFAULTGW=fd22:a075:afd0:e096::1FF
I used a private IPv6 address range,for internal testing.
Node IPv6 routing table.
# ip -6 ro show dev virtbr0
fd22:a075:afd0:e096::/64 proto kernel metric 256 mtu 1500 advmss 1440 hoplimit 4294967295
fe80::/64 proto kernel metric 256 mtu 1500 advmss 1440 hoplimit 4294967295
default via fd22:a075:afd0:e096::1ff metric 1 mtu 1500 advmss 1440 hoplimit 4294967295
#
A test VE container was created,and bound to the Ethernet bridge.
# brctl show
bridge name bridge id STP enabled interfaces
virtbr0 8000.001851a86b76 no eth0
veth6101.0
#
This is the test VE network configuration.
DEVICE=eth0
TYPE=Ethernet
ONBOOT=yes
IPV6INIT=yes
IPV6ADDR=fd22:a075:afd0:e096:65::65/64
From Node,the VE is reachable.
# ping6 -c 3 fd22:a075:afd0:e096:65::65
PING fd22:a075:afd0:e096:65::65(fd22:a075:afd0:e096:65::65) 56 data bytes
64 bytes from fd22:a075:afd0:e096:65::65: icmp_seq=1 ttl=64 time=1.34 ms
64 bytes from fd22:a075:afd0:e096:65::65: icmp_seq=2 ttl=64 time=0.375 ms
64 bytes from fd22:a075:afd0:e096:65::65: icmp_seq=3 ttl=64 time=0.372 ms
--- fd22:a075:afd0:e096:65::65 ping statistics ---
3 packets transmitted, 3 received, 0% packet loss, time 2002ms
rtt min/avg/max/mdev = 0.372/0.696/1.342/0.457 ms
#
From VE,Node is reachable.
# vzctl enter 6101
entered into VE 6101
[root@moe /]# ping6 -c 3 fd22:a075:afd0:e096::101
PING fd22:a075:afd0:e096::101(fd22:a075:afd0:e096::101) 56 data bytes
64 bytes from fd22:a075:afd0:e096::101: icmp_seq=1 ttl=64 time=1.50 ms
64 bytes from fd22:a075:afd0:e096::101: icmp_seq=2 ttl=64 time=0.403 ms
64 bytes from fd22:a075:afd0:e096::101: icmp_seq=3 ttl=64 time=0.401 ms
--- fd22:a075:afd0:e096::101 ping statistics ---
3 packets transmitted, 3 received, 0% packet loss, time 2002ms
rtt min/avg/max/mdev = 0.401/0.769/1.503/0.519 ms
[root@moe /]#
It's pretty straight-forward,and no more worries on link-local breakage.
For IPv4,it was pretty straight-forward too.
This is the IPv4 container's routing.
[root@curly /]# ip ro show
192.168.1.0/24 dev eth0 proto kernel scope link src 192.168.1.64
169.254.0.0/16 dev eth0 scope link
default via 192.168.1.254 dev eth0
This should be self-explainatory.
Ping OpenVZ website.
[root@curly /]# ping -c 3 www.openvz.org
PING www.openvz.org (64.131.90.7) 56(84) bytes of data.
64 bytes from openvz.org (64.131.90.7): icmp_seq=1 ttl=48 time=38.3 ms
64 bytes from openvz.org (64.131.90.7): icmp_seq=2 ttl=48 time=40.1 ms
64 bytes from openvz.org (64.131.90.7): icmp_seq=3 ttl=48 time=38.3 ms
--- www.openvz.org ping statistics ---
3 packets transmitted, 3 received, 0% packet loss, time 2042ms
rtt min/avg/max/mdev = 38.381/38.982/40.184/0.879 ms
[root@curly /]#
What differs here now is,the IPv4 containers are directly routed,and I'm a little concerned about DHCP time-outs,due to X bridged interfaces on source-route.
In using DHCP on the Node for network assignment,required the use of NAT rules,as the IPv4 containers resided on their own subnet.
This simplified route management for a VE.
Since,I have never used a source-route Ethernet bridge setup with DHCP,I think NAT rules still applies,but I will give your aliased interface a shot,and see what happens.
In truth,I'm not a big fan of source-route bridging.
But with IPv6/Ethernet bridging,this is going to be a common practice.
One arena I want to play around with,is using VDE.
I downloaded "openVswitch",compiled,and installed on the test Node.
Time is not a premium for me right now,and my technology mistress is going to accompany me,in finding a good divorce lawyer.(LOL)
|
|
|
|