Hello,
it’s my first post so please be understanding 
I have netdata agent installed in docker in a linux host.
The container is created with SYS_ADMIN and pid:host.
It can see all network interfaces but not the net.eth0 of the host itself. It’s only shown for few secs when starting the agent and then disappears.
I have checked around and made sure the container has the proper permissions. I see in /proc/net/dev inside the container the eth0 is present and has data.
I have another host with same configuration where this issue doesn’t happen. Everything looks identical but…
Any hint what could cause this and how to fix?
Thanks in advance!
Don’t worry, this isn’t Reddit or Stackoverflow, we don’t bite here! 
Can you check if Netdata moved that interface under your docker container? e.g. see the last item on the list of this container here:
We associate veth interfaces with specific containers where possible. That happens with an external script that then corrects the position of the charts. A normal (not virtual) interface shouldn’t be moved though, I’ve never seen that happen.
Well, I see that each container have its own net eth0 item which is related to the virtual interface of the container itself, but there’s no additional net eth0 anywhere around
you can see what I have here:
I wonder if that veth6b6f801 is actually eth0 that hasn’t been renamed/sorted ?
could be because the host is running on a Linode VPS and the eth0 is linked to:
eth0 → …/…/devices/pci0000:00/0000:00:0a.0/virtio8/net/eth0
the “virtio8” is something unexpected?
the other nodes that I have run inside Vmware and the eth0 is linked to /devices/pcixxxx:xx/xxxx:xx:xx.x/net/eth0
no further ideas about this? 
