16.04.1 upgraded server does not start network

[Hammi]

Hi, after upgrade from 14.04, my 16.04 virtual machine does not start it's network:

journalctl -b

shows on eth0

Jul 31 08:01:02 server systemd[1]: Started ifup for eth0.
Jul 31 08:01:02 server sh[1551]: ifup: interface eth0 already configured

However, only lo was brought up, ifconfig shows no trace of eth0.

#ifdown eth0
#ifup eth0

fixes the situation. For a permanent workaround, I can put these two commands into rc.local

Any suggestion for what's going wrong here?

Thanks!

[MAFoElffen]

Moved to Virtualization section for better exposure.

Please post the results of

sudo lshw -class network
ls /sys/class/net
ip -o link show | awk '{print $2,$9}'
cat /proc/net/dev

I think what we will find, is that the device names have changed. If I look at the bootup messages of "upgraded" Ubuntu virts in my console, I notice a message fly by saying something about eth0 being renamed... but that would be just a legacy convention, that might have fell through for you. Wired strictly being named as eth0 (etc.) is now deprecated.

Newer systems now use a different naming system for net devices, other than eth0, eth1... ethX; more like Predictable Network Interface Names, with BIOS input and systemd, now show as something like emX or ensX, and are now variably named based on that auto-gen'ed process. Add in that OS'es now can recognise that they are running as virtualized.

Some people initially fought it, and tried to come up with ways to rename it back to what they were used to... but manually renaming those devices is just more work. If you are just aware that it might have a different name, embrace it, and go with it...

[Hammi]

Thanks for looking into this. If the devices name has changed, why does a manual ifdown eth0 && ifup eth0 help?

In any case, these are the outputs immediately after a boot:

sudo lshw -class network

  *-network DISABLED
       description: Ethernet interface
       product: Virtio network device
       vendor: Red Hat, Inc
       physical id: 3
       bus info: pci@0000:00:03.0
       logical name: eth0
       version: 00
       serial: 52:54:00:64:a4:cf
       width: 32 bits
       clock: 33MHz
       capabilities: msix bus_master cap_list rom ethernet physical
       configuration: autonegotiation=off broadcast=yes driver=virtio_net driverversion=1.0.0 latency=0 link=no multicast=yes
       resources: irq:10 ioport:c060(size=32) memory:febf1000-febf1fff memory:febc0000-febdffff

ls /sys/class/net

eth0
lo

ip -o link show | awk '{print $2,$9}'

lo: UNKNOWN
eth0: DOWN

cat /proc/net/dev

Inter-|   Receive                                                |  Transmit
 face |bytes    packets errs drop fifo frame compressed multicast|bytes    packets errs drop fifo colls carrier compressed
    lo:  166957    2743    0    0    0     0          0         0   166957    2743    0    0    0     0       0          0
  eth0:       0       0    0    0    0     0          0         0        0       0    0    0    0     0       0          0

Thanks again!

[Hammi]

As it turned out, I still had an udev rule 85-ifupdown.rules in /etc/udev/rules.d. Probably a leftover from some old Ubuntu version, this VM exists since 6.06. If I delete this rule, the network comes up reliably.

Thanks for your help!

[MAFoElffen]

You are welcome. Yes yours was named eth0, as your update confirmed. That is just something to check now, as it is "not always" that name now. I could see from other output that your link was down... with you seemes to have fixed by changing your udev rules... Those commands help me see what's going on with virtualized network devices. Even on my own. You can't really see a cable plugged in or a link light on a VM Guest.

Since this is resolved for you now: Please do us a favor and revisit this thread, Link at the upper right of the page "Thread Tools" > Select "Mark Thread as Solved". This help others (later) to search and find answers to their similar problems.