vlan howto

T

tlk_vm

New Member
Jul 21, 2008
12
0
1
Hello,

I'm sure I'm just too blind to see: Where/How do I configure VLANs in the WebUI with beta2?

Thanks a lot in advance.
Regards,
Marc
 
Debian uses a special naming convention for VLANs. Just add the VLAN ID to the ethernet device name, for example "eth0.5" is vlan 5 on eth0.

To add vlan 5 to vmbr0 just add "eth0.5" to the bridge (replacing eth0).

- Dietmar
 
<mode="homer simpson">
Doh!
</mode>
;-)
Ok, I tried to do it the other way around. First create the Interface (that's where I got stuck), then add it to the bridge.

Thank you!

Regards,
Marc
 
When I change eth0 to eth0.5 after a reboot it does not work. I have to issue an
Code: 
 /etc/init.d/networking restart
then is starts to work again.
 
Here is the output of /var/log/messages after a fresh reboot:

Code: 
ACPI: PCI Interrupt 0000:07:00.0[A] -> GSI 16 (level, low) -> IRQ 16
eth0: Broadcom NetXtreme II BCM5708 1000Base-T (B2) PCI-X 64-bit 133MHz found at mem f4000000, IRQ 16, node addr 00:1d:09:6b:db:f9
ACPI: PCI Interrupt 0000:03:00.0[A] -> GSI 16 (level, low) -> IRQ 16
eth1: Broadcom NetXtreme II BCM5708 1000Base-T (B2) PCI-X 64-bit 133MHz found at mem f8000000, IRQ 16, node addr 00:1d:09:6b:db:f7
floppy0: no floppy controllers found
EXT3 FS on dm-1, internal journal
kjournald starting.  Commit interval 5 seconds
EXT3 FS on dm-2, internal journal
EXT3-fs: mounted filesystem with ordered data mode.
kjournald starting.  Commit interval 5 seconds
EXT3 FS on sda1, internal journal
EXT3-fs: mounted filesystem with ordered data mode.
Adding 4194296k swap on /dev/pve/swap.  Priority:-1 extents:1 across:4194296k
Bridge firewalling registered
vmbr0: Dropping NETIF_F_UFO since no NETIF_F_HW_CSUM feature.
802.1Q VLAN Support v1.8 Ben Greear <[EMAIL="greearb@candelatech.com"]greearb@candelatech.com[/EMAIL]>
All bugs added by David S. Miller <[EMAIL="davem@redhat.com"]davem@redhat.com[/EMAIL]>
bnx2: eth0: using MSI
device eth0.6 entered promiscuous mode
audit(1216987780.138:2): dev=eth0.6 prom=256 old_prom=0 auid=4294967295
device eth0 entered promiscuous mode
audit(1216987780.138:3): dev=eth0 prom=256 old_prom=0 auid=4294967295
vmbr0: starting userspace STP failed, starting kernel STP
vmbr1: Dropping NETIF_F_UFO since no NETIF_F_HW_CSUM feature.
bnx2: eth1: using MSI
device eth1.5 entered promiscuous mode
audit(1216987782.674:4): dev=eth1.5 prom=256 old_prom=0 auid=4294967295
device eth1 entered promiscuous mode
audit(1216987782.678:5): dev=eth1 prom=256 old_prom=0 auid=4294967295
vmbr1: starting userspace STP failed, starting kernel STP
bnx2: eth0 NIC Copper Link is Up, 1000 Mbps full duplex
bnx2: eth1 NIC Copper Link is Up, 1000 Mbps full duplex
NET: Registered protocol family 10
ADDRCONF(NETDEV_UP): eth0.6: link is not ready
ADDRCONF(NETDEV_UP): eth1.5: link is not ready
ip_tables: (C) 2000-2006 Netfilter Core Team
tun: Universal TUN/TAP device driver, 1.6
tun: (C) 1999-2004 Max Krasnyansky <[EMAIL="maxk@qualcomm.com"]maxk@qualcomm.com[/EMAIL]>
loaded kvm module (kvm-71)
eth0: no IPv6 routers present
vmbr0: no IPv6 routers present
eth1: no IPv6 routers present
vmbr1: no IPv6 routers present


Then is after the /etc/init.d/networking restart:
Code: 
device eth0.6 left promiscuous mode
audit(1216987863.754:6): dev=eth0.6 prom=0 old_prom=256 auid=4294967295
device eth0 left promiscuous mode
audit(1216987863.754:7): dev=eth0 prom=0 old_prom=256 auid=4294967295
vmbr0: port 1(eth0.6) entering disabled state
device eth1.5 left promiscuous mode
audit(1216987864.094:8): dev=eth1.5 prom=0 old_prom=256 auid=4294967295
device eth1 left promiscuous mode
audit(1216987864.094:9): dev=eth1 prom=0 old_prom=256 auid=4294967295
vmbr1: port 1(eth1.5) entering disabled state
vmbr0: Dropping NETIF_F_UFO since no NETIF_F_HW_CSUM feature.
device eth0.6 entered promiscuous mode
audit(1216987864.266:10): dev=eth0.6 prom=256 old_prom=0 auid=4294967295
device eth0 entered promiscuous mode
audit(1216987864.266:11): dev=eth0 prom=256 old_prom=0 auid=4294967295
vmbr0: starting userspace STP failed, starting kernel STP
vmbr0: port 1(eth0.6) entering listening state
vmbr0: port 1(eth0.6) entering learning state
vmbr0: topology change detected, propagating
vmbr0: port 1(eth0.6) entering forwarding state
vmbr1: Dropping NETIF_F_UFO since no NETIF_F_HW_CSUM feature.
device eth1.5 entered promiscuous mode
audit(1216987865.458:12): dev=eth1.5 prom=256 old_prom=0 auid=4294967295
device eth1 entered promiscuous mode
audit(1216987865.458:13): dev=eth1 prom=256 old_prom=0 auid=4294967295
vmbr1: starting userspace STP failed, starting kernel STP
vmbr1: port 1(eth1.5) entering listening state
vmbr1: port 1(eth1.5) entering learning state
vmbr1: topology change detected, propagating
vmbr1: port 1(eth1.5) entering forwarding state
eth0.6: no IPv6 routers present
vmbr0: no IPv6 routers present
vmbr1: no IPv6 routers present
eth1.5: no IPv6 routers present


This is my /etc/network/interface:

Code: 
auto lo
iface lo inet loopback
auto vmbr0
iface vmbr0 inet static
        address 10.4.100.66
        netmask 255.255.255.0
        gateway 10.4.100.1
        bridge_ports eth0.6
        bridge_stp on
        bridge_fd 0
auto vmbr1
iface vmbr1 inet manual
        bridge_ports eth1.5
        bridge_stp on
        bridge_fd 0

If I do not use a VLAN then it works fine after a reboot.
 
Hello,

confirmed. Tagged interfaces work only after restarting network stack. While playing around with it, I found out the following so far:

eth0 untagged, no bridge: works.
eth0 in bridge: works.

eth0.x used without bridge: works
eth0.x used in bridge: sends packet _untagged_ through eth0
If you now restart the network stack it mostly works, but not always. (8 out of 10 times)

Not usable at all. :(

If time permits, I'll try that with an unpatched kernel, just to make sure it's not a basic 802.1Q problem with the bridge.
But this will be no sooner then next week: out of office.

Just to make it clear: Apart from that the VE is just great. Thinking of deploying it in a production environment.

Regards,
Marc
 
Did I fall off my rocker, or shouldn't I be able to specify a VLAN per VE/KVM?

For example, if my network is setup in a way that my cluster nodes are plugged into TRUNK ports, how ould VE100 be on vlan 5 and VE101 be on vlan 6?

Am I expecting too much, or just not understanding this?
 
Let me attempt to redeem my self.

So I setup the interfaces on the NODES before creating the VE??

For example, if I make a new bridge device, vmbr1, I should specify eth0.5 as the bridged interfaces? Correct?
 
zccopwrx said:
For example, if my network is setup in a way that my cluster nodes are plugged into TRUNK ports, how ould VE100 be on vlan 5 and VE101 be on vlan 6?
Click to expand...

Just connect vmbr0 directly to eth0, and configure the vlan inside the VM.

or

connect vmbr0 to vlan5 (eth0.5), and vmbr1 to vlan6 (eth0.6)

I don't have a vlan setup here, so I cant test.

- Dietmar
 
vlan

dietmar said:
Debian uses a special naming convention for VLANs. Just add the VLAN ID to the ethernet device name, for example "eth0.5" is vlan 5 on eth0.

To add vlan 5 to vmbr0 just add "eth0.5" to the bridge (replacing eth0).

- Dietmar
Click to expand...


but, do you have any plan for affing a new interface (vlan) to the machine
via webUI?

Ciao, Diaolin
 
You can use eth0.XXX on the web interface too (at least for bridge slaves). Or what do you want exactly?

- Dietmar
 
re: vlan

dietmar said:
You can use eth0.XXX on the web interface too (at least for bridge slaves). Or what do you want exactly?

- Dietmar
Click to expand...

OK, but adding eth0.XXX does create the additional vlan for me or should
i add the vlan into the /etc/network/interfaces?

I did not test this solution and i would like if the interface does the work
:-)

Diaolin
 
vlan...

dietmar said:
should be possible.



Because you want to duplicate network traffic?
Click to expand...


I need to have eth0 for vmbr0
eth0.50 for vmbr1
eth0.51 for vmbr2

Associating the vlans on the switcher i can have multiple
vlan on the same interface.

OK, this is not the best solution but is the only way i can "separate"
the bridges on the same interface.

On the eth0 i can dump all, of course.
But i hope that the Proxmox admin should verify the traffic.

Diaolin
 
vlan the Debian way

dietmar said:
Ah, yes - I think that makes sense.

- Dietmar
Click to expand...

If we want to use a vlan on Debian the best mode is this:

iface eth0 inet static
address xxx.xxx.xxx.xx
netmask xxx.xxx.xxx.xxx

iface vlan1 inet static
vlan-raw-device eth0
address 192.168.1.1
netmask 255.255.255.0

If you implement a method in the interface for reading the
/etc/network/interfaces intercepting the vlan-raw-device
it can be simple to write the "vlan create" button.

In few words: we have "Add bond... Add bridge"
why not have "Add vlan..."
Like the add bridge with
Vlan name vlan[{INPUT}number from 1 to 99]
IP ADDRESS []
NETMASK []
START on boot
REAL INTERFACE [choose between existing interfaces (dropdown)]

Is it possible?

:-)

The real result is that adding vlan to the vmbr1 lets all the machines (tap) connected to it
to work transparently withouth tagging. The tags are on the OUT-interface and on the switch.
The virtuals are all normal withouth vlan's.
Simple, no?

:-)

I think that this can be very useful

Tx, diaolin
 
Last edited:
diaolin said:
The real result is that adding vlan to the vmbr1 lets all the machines (tap) connected to it
to work transparently withouth tagging. The tags are on the OUT-interface and on the switch.
The virtuals are all normal withouth vlan's.
Simple, no?
Click to expand...

Yes, and that works already (just use ethX.YYY).
 
You must log in or register to reply here.
Top Bottom