Re: Kernel protocol and different namespaces

[Maria Matejka]

On 6/7/19 12:14 PM, Maria Jan Matějka wrote:

Thinking once more about it, with respect to the interfaces and so, the BGP transported over Unix socket seems to be quite simple feature to do.

# netstat -aFunix
Kernel Interface table

eno1      1500 77564888      0    614 0      66111123      0      0 0 BMRU
lo       65536 68143909      0      0 0      68143909      0      0 0 LRU
# ip netns add test
# ip netns exec test /bin/netstat -aFunix
Kernel Interface table

# ip netns del test

So, I'm not sure if that's going to work the way that you want.

I thought, and my initial tests support, that Unix sockets are network namespace specific. Iface      MTU    RX-OK RX-ERR RX-DRP RX-OVR    TX-OK TX-ERR TX-DRP TX-OVR Flg Iface      MTU    RX-OK RX-ERR RX-DRP RX-OVR    TX-OK TX-ERR TX-DRP TX-OVR Flg lo       65536        0      0      0 0             0      0      0 0 L

It will work the same way as the BIRD control socket works.


Maria

You can try it by the attached script (run by root) which uses socat for demonstration.

socat-test.sh
Description: application/shellscript