Routed notwork is not working with Centos 7 now [from the box]

I got similar to Nictype=routed: network is not working with Ubuntu 18.04 (but ok with Centos 7) issue with centos 7 now on LXD 4.1 from snap:

lxc launch images:${ostemplate} ${ct}

lxc stop ${ct}
lxc config device add ${ct} root disk path=/ pool=lxd size=${disk}
lxc config device add $ct eth0 nic nictype=routed parent=eno1 ipv4.address=$ip
lxc start ${ct}

Inside ct:

# ip a
1: lo: <LOOPBACK,UP,LOWER_UP> mtu 65536 qdisc noqueue state UNKNOWN group default qlen 1000
    link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
    inet scope host lo
       valid_lft forever preferred_lft forever
    inet6 ::1/128 scope host
       valid_lft forever preferred_lft forever
2: eth0@if31: <BROADCAST,MULTICAST> mtu 1500 qdisc noqueue state DOWN group default qlen 1000
    link/ether 22:3e:50:77:3d:9e brd ff:ff:ff:ff:ff:ff link-netnsid 0
    inet 194.117.***.***/32 brd scope global eth0
       valid_lft forever preferred_lft forever

and network is not working, I’m unable to ping even

I see:

# ps auxfw
root       223  0.0  0.0 115524  2192 ?        Ss   00:01   0:00 bash
root       232  0.0  0.0 155448  2048 ?        R+   00:01   0:00  \_ ps auxfw
root         1  0.0  0.0  43284  3468 ?        Ss   00:01   0:00 /sbin/init
root        38  0.0  0.0  39092  5368 ?        Ss   00:01   0:00 /usr/lib/systemd/systemd-journald
root        51  0.0  0.0  41536  1908 ?        Ss   00:01   0:00 /usr/lib/systemd/systemd-udevd
root        59  0.0  0.0  12100  2172 ?        Ss   00:01   0:00 /bin/bash /etc/rc.d/init.d/network start
root       183  0.0  0.0  12096  2292 ?        S    00:01   0:00  \_ /bin/bash /etc/sysconfig/network-scripts/ifup-eth ifcfg-eth0.deleted boot
root       199  0.0  0.0 102908  5812 ?        S    00:01   0:00      \_ /sbin/dhclient -1 -q -lf /var/lib/dhclient/ -pf /var/run/dhcl
root        61  0.0  0.0  24268  1756 ?        Ss   00:01   0:00 /usr/lib/systemd/systemd-logind
dbus        63  0.0  0.0  58012  2536 ?        Ss   00:01   0:00 /usr/bin/dbus-daemon --system --address=systemd: --nofork --nopidfile --systemd-activatio
root        71  0.0  0.0  22704  1808 ?        Ss   00:01   0:00 /usr/sbin/crond -n
root        76  0.0  0.0   6524  1116 console  Ss+  00:01   0:00 /sbin/agetty --noclear --keep-baud console 115200,38400,9600 linux

so it looks like some problem with dhclient, but unfortunately I was not able to force it not touching network like I did to Ubuntu.

Does anybody have this issue? How to fix it? Thanks

Take a look at the config files in /etc/sysconfig/network-scripts/ as that is where they are stored for CentOS and your ps output shows thats what is running dhclient.

It looks like its still trying to process ifcfg-eth0.deleted so suggest removing that file entirely.

Hi. @Nick_Knutov and @tomp

I also tested it.
It doesn’t work with centos 7 containers, but works fine with centos 8 containers.

@tomp, 2 questions.

  1. When adding a nic device, if i add the device name as eth2, if eth1 does not exist in the container, is it automatically registered as eth1 in the container?
  2. Is routed nic not available in vm?

Thank you.

Seems to work OK, here’s my process:

lxc init images:centos/7 c1
lxc config device add c1 eth0 nic nictype=routed ipv4.address= parent=enp3s0
lxc start

lxc ls c1
| NAME |  STATE  |         IPV4         | IPV6 |   TYPE    | SNAPSHOTS |
| c1   | RUNNING | (eth0) |      | CONTAINER | 0         |

OK good, so container has IP without me doing anything.

But lets check connectivity:

ing -c2 -W1
PING ( 56(84) bytes of data.

--- ping statistics ---
2 packets transmitted, 0 received, 100% packet loss, time 1006ms

Oh dear, something isn’t right. Lets take a look at the container’s routes:

lxc exec c1 -- ip r

OK, no output, this means no default route. This is our problem.

lxc exec c1 -- cat /etc/sysconfig/network-scripts/ifcfg-eth0

Change BOOTPROTO=dhcp to BOOTPROTO=none.

lxc restart c1
ping -c2 -W1
PING ( 56(84) bytes of data.
64 bytes from icmp_seq=1 ttl=64 time=0.403 ms
64 bytes from icmp_seq=2 ttl=64 time=0.065 ms

--- ping statistics ---
2 packets transmitted, 2 received, 0% packet loss, time 1022ms
rtt min/avg/max/mdev = 0.065/0.234/0.403/0.169 ms

Working :slight_smile:

If you want to be sure what the device name will be inside the container, be sure to specify name in the device config, e.g.

lxc config device add c1 mynic nic nictype=routed name=eth2 ipv4.address= parent=enp3s0

Ah, but the routing still isn’t quite right:

lxc exec c1 -- ip r dev eth0 scope link metric 1002 

That gets us to the default gateway, and thus the LXD host, but not onto the wider network.

Try adding this to the container as well as the change above.

lxc exec c1 -- cat /etc/sysconfig/network-scripts/route-eth0 dev eth0
default via dev eth0
1 Like

Thank you @tomp

Please take a look at my container’s configuration first.


What is wrong with my configuration?

See my earlier reply.

Also, which device are you wanting to be your default gateway device (assuming the static route gets added correctly, you cant have lxdbr0 provide a default route by DHCP and have one defined statically).

So aside from the particularities of CentOS 7 network setup, this configuration is unlikely to work properly as you have two competing NICs.

Thank you @tomp.

As you said, if i add routed nic as eth0 and write route-eth0, it will ping from the container to the outside (

However, adding routed-nic as eth1 doesn’t seem to work.
As you can see in the picture above, ifcfg-eth1 is also written and route-eth1 is not working.

However, in case of centos 8, by default, only routed-nic is added, and external communication is good.

Please refer to the picture below.

I’m not clear what you are trying to achieve.

Your container config shows two NICs; eth0 and mynic, which will appear as eth0 and eth1 respectively inside the container.

Which one of those should be the default gateway device?

Thank you @tomp.

Add routed-nic in centos 7 and 8 and look at netstat -an as below.

After creating /etc/sysconfig/network-scripts/route-eth1 in centos 7 and rebooting, does not register as the default gateway.

So, i register as the default gw with the route add command in centos7, external communication is good.

Thanks again for letting me know in detail.

1 Like