Ask Your Question

Revision history [back]

click to hide/show revision 1
initial version

No VXLAN Traffic Across Compute Nodes

I have an environment deployed using OpenStack-Ansible. Three control plane nodes and three compute nodes, using VXLAN and Linux bridge (not OVS).

For the life of me, I cannot figure out why I can't get pings going between instances on two different compute nodes, which I think I've narrowed down to being a problem with VXLAN.

Instances on the same node can ping just fine, after I set static IPs (assuming DHCP & metadata isn't working because of the same issue).

Running tcpdump on the instance tap interface, all the way down to the VLAN-tagged interface attached to br-vxlan, I see the ARP requests (encapsulated further down), but no replies. I also don't see them on any other host from the same interface.

21:22:19.347214 IP > OTV, flags [I] (0x08), overlay 0, instance 15 ARP, Request who-has tell, length 28

The way I understand the packet flow is this:

tapXXXXXXXX-XX --> brqXXXXXXXX-XX --> vxlan-xx --> em1.1723

Initially, my issue was that my network team didn't have multicast enabled for my tunnel VLAN. Since then, they've turned it on (sparse mode, if that matters) and I can join the multicast group on each machine using iperf and send a multicast message from any of the six machines and receive it on all of them, so I know basic multicast is working, but I'm wondering if something is still not configured quite right?

I'm lost as to where to look next. I suck at multicast, beyond the basics and I'm pretty new to VXLAN as well.

Any suggestions would be greatly appreciated. Let me know if more information is helpful.