I found a typo in my rspec. I fixed it, resubmitted, and after a few
false starts, the shared vlan and stitching bits are working. This is
great!
One issue remains that I'm hoping someone could look into. My
experimental configuration is such that I would like to have all my
interfaces in the same IP subnet. So I have something like this:
BBN (shared_vlan) - (bbn bridge VM) - circuit - (utah bridge VM) - Utah
(shared_vlan)
On the bridge VMs, I bridge the two interfaces. For those interested:
brctl addbr br0
brctl addif br0 eth1
brctl addif br0 eth2
ifconfig br0 up
(then optionally assign IP to br0, flush IPs on eth1/eth2)
So now from Utah bridge VM I can ping the pre-created BBN VMs in the
shared vlan over the circuit all in the same IP subnet. Perfect.
However, I did the exact same thing on the PG Utah side but I can't ping
any of the pre-create Utah VMs, from either the utah bridge or the BBN side.
I double-checked the POA for pg_utah, it succeeded. The two Utah VMs
can talk to each other, just not the other slice(s). If there a way to
verify the Utah switch vlan config?
The Utah VMs in question are:
pcvm480-1.emulab.net (bridge)
pcvm218-1.emulab.net (shared vlan VM0)
pcvm238-1.emulab.net (shared vlan VM1)
thanks,
- ezra