VLAN connectivity issues on vRACK between my dedicated servers

Bare Metal and VPS
1

Has anyone had trouble with VLAN connectivity over vRACK?

Sending server (Server #1):

# arping 192.168.69.2 -c 1
ARPING 192.168.69.2 from 192.168.69.1 vmbr1.10
Sent 1 probes (1 broadcast(s))
Received 0 response(s)

Packet is sent, but the reply packet from Server #2 is missing the VLAN 10 header.

# sudo tcpdump -i enp8s0f1np1 -e -n -v
tcpdump: listening on enp8s0f1np1, link-type EN10MB (Ethernet), snapshot length 262144 bytes
23:37:44.531281 d8:43:ae:8a:6f:5f > ff:ff:ff:ff:ff:ff, ethertype 802.1Q (0x8100), length 46: vlan 10, p 0, ethertype ARP (0x0806), Ethernet (len 6), IPv4 (len 4), Request who-has 192.168.69.2 (ff:ff:ff:ff:ff:ff) tell 192.168.69.1, length 28
23:37:44.531393 9c:6b:00:67:06:c2 > d8:43:ae:8a:6f:5f, ethertype ARP (0x0806), length 60: Ethernet (len 6), IPv4 (len 4), Reply 192.168.69.2 is-at 9c:6b:00:67:06:c2, length 46

Issue: Reply packet is stripped of the VLAN 10 header

Receiving server (Server #2):

# sudo tcpdump -i enp1s0f1np1 -e -n -v
tcpdump: listening on enp1s0f1np1, link-type EN10MB (Ethernet), snapshot length 262144 bytes
10:09:14.348946 d8:43:ae:8a:6f:5f > ff:ff:ff:ff:ff:ff, ethertype 802.1Q (0x8100), length 60: vlan 10, p 0, ethertype ARP (0x0806), Ethernet (len 6), IPv4 (len 4), Request who-has 192.168.69.2 (ff:ff:ff:ff:ff:ff) tell 192.168.69.1, length 42
10:09:14.348959 9c:6b:00:67:06:c2 > d8:43:ae:8a:6f:5f, ethertype 802.1Q (0x8100), length 46: vlan 10, p 0, ethertype ARP (0x0806), Ethernet (len 6), IPv4 (len 4), Reply 192.168.69.2 is-at 9c:6b:00:67:06:c2, length 28

Reply packet leaves server #2 with VLAN 10 header tag

So between Server #2 and Server #1, the reply packet is being stripped of VLAN header; either by the vRACK switches or possibly the network card itself. Server #1 is an Intel NIC while Server #2 is Broadcom. My network configuration appears correct as both servers are sending tagged packets out. Only Server #1 doesn't receive tagged packets, they come in untagged from the NIC. Server #2 receives tagged packets consistently.

Has anyone seen this issue before? I have had no luck with getting OVHCloud support to assist with this and I'm very much exhausted all other options in diagnosing the issue. For context I am running Proxmox 8 which is Debian based.

Thanks very much

2

Hi,

I'm unable to find your servers with the provided MAC addresses. Are these Advance servers? I know this range has Intel E810 and Broadcom BCM57502 NICs.
Could you please share the server IDs or support ticket number? I'll try to reproduce the issue. Does it also happen in rescue mode?

3

Yes these are Advanced servers:


499195 - Intel NIC (this appears to be the problematic one - sends with VLAN tag but receives replies without)

492185 - Broadcom NIC (sends/receives packets with VLAN tag)
Ticket number is: CS10879494

Thanks very much! I haven't been able to use my new server due to this issue as I require VLAN connectivity.