List, I am having trouble setting up Xen with VLANs. I am only using one network interface, and I plan to use this interface for all of the guest instances as well. I have the switch port configured as a trunk where it will receive the traffic for multiple VLANs. That said, the host that I am configuring Xen on has network connectivity when I run ''vconfig add eth0 8'' and I use the interface eth0.8. Now I have some basic questions. How is veth0 connected to vif0.0? Sorry if it is an obvious answer. Also, I understand that I should have one bridge for each VLAN and that for each domU I can connect to the respective bridge/VLAN. I cant even get dom0 to work on the VLAN mentioned above behind a bridge. I suspect the configuration I am looking for looks like: eth0->vif0.0->xenbr8->peth0.8->peth0 Where the only device configured with a MAC and IP is eth0. Is this correct? Is the end result possible to achieve with the provided ''network-bridge'' script? Any help will be much appreciated. _______________________________________________ Xen-users mailing list Xen-users@lists.xensource.com http://lists.xensource.com/xen-users
On Wednesday 08 March 2006 5:06 pm, Brian Wong wrote:> List, > I am having trouble setting up Xen with VLANs. I am only using one > network interface, and I plan to use this interface for all of the > guest instances as well. I have the switch port configured as a trunk > where it will receive the traffic for multiple VLANs. That said, the > host that I am configuring Xen on has network connectivity when I run > ''vconfig add eth0 8'' and I use the interface eth0.8.have you used 802.1Q before? there are some quirks with the MTU. in my case, i had to put eth0 to 1504, and since unpatched 100baseT drivers didn''t allow more than 1500, i had to use 1000baseT, even if i didn''t need the speed. but i think you want the software bridge to move ''tagged'' packets, so it had to support MTU 1504 too. i''m not sure it works. also, the vif drivers would have to support MTU 1504.... there i think are some bad things... i think it''s easier to ''untag'' the VLANs in dom0, so you would have several eth0.x, and then put those in several bridges, connecting to the appropriate vif''s; just like having several physical eth cards. for that to work, you''d have to use the physical eth0, not the ''virtualized'' eth0 you have after the Xen scripts do their magic. -- Javier _______________________________________________ Xen-users mailing list Xen-users@lists.xensource.com http://lists.xensource.com/xen-users
On Wed, Mar 08, 2006 at 05:23:34PM -0500, Javier Guerra wrote: [snip]> i think it''s easier to ''untag'' the VLANs in dom0, so you would have several > eth0.x, and then put those in several bridges, connecting to the appropriate > vif''s; just like having several physical eth cards. > > for that to work, you''d have to use the physical eth0, not the ''virtualized'' > eth0 you have after the Xen scripts do their magic.Hello, I have tried to add an eth0.x interface into a xen bridge using the network-bridge script, but could not get it running. Have you (or anyone else on the list) actually managed to do something like that? Matus _______________________________________________ Xen-users mailing list Xen-users@lists.xensource.com http://lists.xensource.com/xen-users
On 3/8/06, Matus Harvan <m.harvan@iu-bremen.de> wrote:> On Wed, Mar 08, 2006 at 05:23:34PM -0500, Javier Guerra wrote: > [snip] > > i think it''s easier to ''untag'' the VLANs in dom0, so you would have several > > eth0.x, and then put those in several bridges, connecting to the appropriate > > vif''s; just like having several physical eth cards. > > > > for that to work, you''d have to use the physical eth0, not the ''virtualized'' > > eth0 you have after the Xen scripts do their magic. > > Hello, > > I have tried to add an eth0.x interface into a xen bridge using the > network-bridge script, but could not get it running. Have you (or anyone > else on the list) actually managed to do something like that? >Ive been tinkering with it all day and havent been able to get it to work. This doesnt seem to work because the script brings down the ''netdev'' and when that device is ''eth0.X'' where X is the vlan, it totally removes it. Then the script later tries to rename it to peth0.X but the ''eth0.X'' no longer exists. _______________________________________________ Xen-users mailing list Xen-users@lists.xensource.com http://lists.xensource.com/xen-users
On Wednesday 08 March 2006 5:38 pm, Brian Wong wrote:> This doesnt seem to work because the script brings down the ''netdev'' > and when that device is ''eth0.X'' where X is the vlan, it totally > removes it. Then the script later tries to rename it to peth0.X but > the ''eth0.X'' no longer exists.i believe (can be wrong) that the xen scripts have two main parts: one for initial setup, and one for adding each domU to the right bridges. try to totally disable the initial setup script. don''t let it confuse anything. then write your own script that sets all the VLAN devices, and all needed bridges. try to make it set all you need, ready to receive the domU''s vifs. then the second part of the Xen scripts should work with little or no modifications. after you get it working, it might be wise to add some of the usual xen renaming scheme to your initial script, and replace the Xen provided with yours. -- Javier _______________________________________________ Xen-users mailing list Xen-users@lists.xensource.com http://lists.xensource.com/xen-users
Hi It does not work. You can get the VLAN tagging working with standard network sripts. Greg has made some good examples. http://lists.xensource.com/archives/html/xen-devel/2005-10/msg01174.html http://lists.xensource.com/archives/html/xen-devel/2006-01/msg00562.html However the traffic does not go through without the patch for the ethtool. I hope that someone would put the modifications to standard package. I never managed to create rpm-package with required modifications. -- Timo ----- Alkuperäinen viesti ----- From: Javier Guerra <javier@guerrag.com> To: xen-users@lists.xensource.com Lähetetyt: 9. maaliskuuta 2006 0:45:14 GMT+0200 Aihe: Re: [Xen-users] Bridging and VLANs 802.1Q trunk On Wednesday 08 March 2006 5:38 pm, Brian Wong wrote:> This doesnt seem to work because the script brings down the ''netdev'' > and when that device is ''eth0.X'' where X is the vlan, it totally > removes it. Then the script later tries to rename it to peth0.X but > the ''eth0.X'' no longer exists.i believe (can be wrong) that the xen scripts have two main parts: one for initial setup, and one for adding each domU to the right bridges. try to totally disable the initial setup script. don''t let it confuse anything. then write your own script that sets all the VLAN devices, and all needed bridges. try to make it set all you need, ready to receive the domU''s vifs. then the second part of the Xen scripts should work with little or no modifications. after you get it working, it might be wise to add some of the usual xen renaming scheme to your initial script, and replace the Xen provided with yours. -- Javier _______________________________________________ Xen-users mailing list Xen-users@lists.xensource.com http://lists.xensource.com/xen-users
On 3/8/06, Timo Virtaneva <timo@virtaneva.fi> wrote:> Hi > > It does not work. You can get the VLAN tagging working with standard network sripts. Greg has made some good examples. > > > http://lists.xensource.com/archives/html/xen-devel/2005-10/msg01174.html > http://lists.xensource.com/archives/html/xen-devel/2006-01/msg00562.html > > However the traffic does not go through without the patch for the ethtool. I hope that someone would put the modifications to standard package. I never managed to create rpm-package with required modifications. > > -- Timo >How can I diagnose whether I am affected by either the MTU issue or the ethtool checksum problem? Here is my ''ip addr'' and ''brctl show''. I still cant figure out why I have no connectivity. Is there something obviously wrong with my config? 1: eth0: <BROADCAST,MULTICAST,UP> mtu 1500 qdisc pfifo_fast qlen 1000 link/ether fe:ff:ff:ff:ff:ff brd ff:ff:ff:ff:ff:ff 2: eth1: <BROADCAST,MULTICAST,UP> mtu 1500 qdisc pfifo_fast qlen 1000 link/ether 00:0e:7f:25:8f:ae brd ff:ff:ff:ff:ff:ff inet 10.0.0.42/8 brd 10.255.255.255 scope global eth1 3: lo: <LOOPBACK,UP> mtu 16436 qdisc noqueue link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00 inet 127.0.0.1/8 scope host lo 4: vif0.0: <BROADCAST,MULTICAST,UP> mtu 1500 qdisc noqueue link/ether fe:ff:ff:ff:ff:ff brd ff:ff:ff:ff:ff:ff 5: veth0: <BROADCAST,MULTICAST,UP> mtu 1500 qdisc noqueue link/ether 00:0e:7f:25:8f:af brd ff:ff:ff:ff:ff:ff inet 172.16.8.80/24 brd 172.16.255.255 scope global veth0 6: vif0.1: <BROADCAST,MULTICAST> mtu 1500 qdisc noop link/ether fe:ff:ff:ff:ff:ff brd ff:ff:ff:ff:ff:ff 7: veth1: <BROADCAST,MULTICAST> mtu 1500 qdisc noop link/ether 00:00:00:00:00:00 brd ff:ff:ff:ff:ff:ff 8: vif0.2: <BROADCAST,MULTICAST> mtu 1500 qdisc noop link/ether fe:ff:ff:ff:ff:ff brd ff:ff:ff:ff:ff:ff 9: veth2: <BROADCAST,MULTICAST> mtu 1500 qdisc noop link/ether 00:00:00:00:00:00 brd ff:ff:ff:ff:ff:ff 10: vif0.3: <BROADCAST,MULTICAST> mtu 1500 qdisc noop link/ether fe:ff:ff:ff:ff:ff brd ff:ff:ff:ff:ff:ff 11: veth3: <BROADCAST,MULTICAST> mtu 1500 qdisc noop link/ether 00:00:00:00:00:00 brd ff:ff:ff:ff:ff:ff 12: vif0.4: <BROADCAST,MULTICAST> mtu 1500 qdisc noop link/ether fe:ff:ff:ff:ff:ff brd ff:ff:ff:ff:ff:ff 13: veth4: <BROADCAST,MULTICAST> mtu 1500 qdisc noop link/ether 00:00:00:00:00:00 brd ff:ff:ff:ff:ff:ff 14: vif0.5: <BROADCAST,MULTICAST> mtu 1500 qdisc noop link/ether fe:ff:ff:ff:ff:ff brd ff:ff:ff:ff:ff:ff 15: veth5: <BROADCAST,MULTICAST> mtu 1500 qdisc noop link/ether 00:00:00:00:00:00 brd ff:ff:ff:ff:ff:ff 16: vif0.6: <BROADCAST,MULTICAST> mtu 1500 qdisc noop link/ether fe:ff:ff:ff:ff:ff brd ff:ff:ff:ff:ff:ff 17: veth6: <BROADCAST,MULTICAST> mtu 1500 qdisc noop link/ether 00:00:00:00:00:00 brd ff:ff:ff:ff:ff:ff 18: vif0.7: <BROADCAST,MULTICAST> mtu 1500 qdisc noop link/ether fe:ff:ff:ff:ff:ff brd ff:ff:ff:ff:ff:ff 19: veth7: <BROADCAST,MULTICAST> mtu 1500 qdisc noop link/ether 00:00:00:00:00:00 brd ff:ff:ff:ff:ff:ff 20: eth0.8: <BROADCAST,MULTICAST,PROMISC,UP> mtu 1500 qdisc noqueue link/ether fe:ff:ff:ff:ff:ff brd ff:ff:ff:ff:ff:ff 21: xenbr8: <BROADCAST,MULTICAST,UP> mtu 1500 qdisc noqueue link/ether fe:ff:ff:ff:ff:ff brd ff:ff:ff:ff:ff:ff bridge name bridge id STP enabled interfaces xenbr8 8000.feffffffffff no eth0.8 vif0.0 _______________________________________________ Xen-users mailing list Xen-users@lists.xensource.com http://lists.xensource.com/xen-users