Quoting Ben Pfaff ([email protected]): > On Tue, Jun 16, 2015 at 11:43:17PM +0200, Stefan Kooman wrote: -- snip-- >>How can one convert a trunk port to an "access port" > > ovs-vsctl set port tap0 tag=9 -- clear port tap0 trunks
Thanks, this works like a charm.
I would like to understand what functionality a mixed "tagged" and
"trunk" port provide? As far as I can see (tcpdump) the traffic bound for the
vnet (tap port) never reaches it's destination anymore after being
tagged (although being a trunk port as well). Shouldn't these modes be mutually
exclusive? I'm running an experimental "nested" virtualization solution:
hypervisor config (L0):
Port "vnet6"
tag: 226
trunks: [226, 227]
Interface "vnet6"
guest hypervisor config (L1):
Bridge uplink
Port uplink
Interface uplink
type: internal
Port "eth1"
trunks: [226, 227]
Interface "eth1"
Port "vnet0"
tag: 226
Interface "vnet0"
So, eth1 here (L1) is actually vnet6 on hypervisor (L0). If vnet6 (L0) has both
trunk and tag the traffic never reaches eth1 (L1) anymore (and of course also
not vnet0: guest vm (L2)).
What I would expect (but does not seem to happen) is to see traffic with vlan
tag "226"
be both passed tagged as wel as untagged for vnet6 (L0).
Does this make sense?
Thanks,
Stefan Kooman
--
| BIT BV http://www.bit.nl/ Kamer van Koophandel 09090351
| GPG: 0xD14839C6 +31 318 648 688 / [email protected]
pgpcUQO3JLDCq.pgp
Description: PGP signature
_______________________________________________ discuss mailing list [email protected] http://openvswitch.org/mailman/listinfo/discuss
