1

I have a Debian Jessie 8.2 server serving as an Xen VM host. The server has two physical network interfaces that are bonded together using LACP. There is also an xenbr0 interface. I would like to enable jumbo frames, however I am unable to. Here's the network configuration:

auto eth0
iface eth0 inet manual
   bondmaster bond0
   mtu 9000

auto eth1
iface eth1 inet manual
   bondmaster bond0
   mtu 9000

auto bond0
iface bond0 inet manual
   bond-mode 4
   bond-miimon 100
   bond-lacp-rate 1
   bond-slaves eth0 eth1
   mtu 9000

auto xenbr0
iface xenbr0 inet static
   bridge_ports bond0
   address 10.0.0.12
   gateway 10.0.0.1
   netmask 255.255.255.0
   mtu 9000

After adding MTU to each interface, I ran sudo ifdown bond0 && sudo ifup bond0. The MTU is still reported as 1,500. I then tried the same with the xenbr0 interface, however I get the following output:

user@host:~$ sudo ifdown xenbr0 && sudo ifup xenbr0
ifdown: interface xenbr0 not configured

Waiting for xenbr0 to get ready (MAXWAIT is 32 seconds).
RTNETLINK answers: File exists
Failed to bring up xenbr0.

Here's the output from ip a, without virtual interfaces for VM guests:

1: lo: <LOOPBACK,UP,LOWER_UP> mtu 65536 qdisc noqueue state UNKNOWN group default 
    link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
    inet 127.0.0.1/8 scope host lo
       valid_lft forever preferred_lft forever
    inet6 ::1/128 scope host 
       valid_lft forever preferred_lft forever
2: eth0: <BROADCAST,MULTICAST,SLAVE,UP,LOWER_UP> mtu 1500 qdisc pfifo_fast master bond0 state UP group default qlen 1000
    link/ether e8:94:f6:08:aa:30 brd ff:ff:ff:ff:ff:ff
3: eth1: <BROADCAST,MULTICAST,SLAVE,UP,LOWER_UP> mtu 1500 qdisc pfifo_fast master bond0 state UP group default qlen 1000
    link/ether e8:94:f6:08:aa:30 brd ff:ff:ff:ff:ff:ff
4: bond0: <BROADCAST,MULTICAST,MASTER,UP,LOWER_UP> mtu 1500 qdisc noqueue master xenbr0 state UP group default 
    link/ether e8:94:f6:08:aa:30 brd ff:ff:ff:ff:ff:ff
17: xenbr0: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdisc noqueue state UP group default 
    link/ether e8:94:f6:08:aa:30 brd ff:ff:ff:ff:ff:ff
    inet 10.0.0.12/24 brd 10.0.0.255 scope global xenbr0
       valid_lft forever preferred_lft forever
    inet6 2001:****:****:****:****:****:****:****/64 scope global mngtmpaddr dynamic 
       valid_lft 2591736sec preferred_lft 604536sec
    inet6 fe80::ea94:f6ff:fe08:aa30/64 scope link 
       valid_lft forever preferred_lft forever

What have I done wrong?

  • After a little research, it seems the onboard NIC does not support jumbo frames. Using ip link set dev <dev> mtu 9000 on every interface however works, including apparently the onboard NIC. Looking at a packet capture, I can see jumbo frames are indeed being used over the bond interface. I'm not sure exactly what's going on though, since the manufacturer's site specifically states that the NIC does not have support for jumbo frames. Anyway. After a reboot, the MTU is back to 1,500 and not read from the config. Perhaps the MTU is being set in the wrong order? – Torvero Dec 20 '15 at 18:06

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Browse other questions tagged or ask your own question.