Yes, its to be expected. IIRC, you were helping me investigate the same thing at the TripleO mid-cycle. :) The port gets the necessary configuration setup by the DHCP agent to take care of PXE, but this being baremetal there is no hypervisor on which some agent is running and plugging VIFs into a vSwitch ports at the compute host level. We rely (at least in devstack) on a flat networking environment setup where baremetal nodes are already tapped into the tenant network namespace without the help of the agents. I believe devtest sets up something similar on the host thats running the devtest VMs. The end result is ports showing DOWN tho they are functional.
On Wed, Aug 27, 2014 at 5:06 PM, Carlino, Chuck <chuck.carl...@hp.com> wrote: > Hi all, > > I'm working on bug [1] and have code in review[2]. The bug wants > neutron's port['mac_address'] to be updatable to handle the case where an > ironic server's nic is replaced. A comment in the review suggests that we > only allow mac address update when the port status is not ACTIVE. While > I've made another change which may (or may not) address the underlying > concern, I want to ask if port status is correct for ironic server ports. > When I run devtest (tripleo) on my laptop, I get VM ironic servers, which > when booted all have neutron ports with 'binding:vif_type': > 'binding_failed' and 'status': 'DOWN'. Is this expected? Does this happen > with hardware ironic servers? > > Thanks in advance! > Chuck > > [1] https://bugs.launchpad.net/neutron/+bug/1341268 > [2] https://review.openstack.org/112129 > > > _______________________________________________ > OpenStack-dev mailing list > OpenStack-dev@lists.openstack.org > http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack-dev >
_______________________________________________ OpenStack-dev mailing list OpenStack-dev@lists.openstack.org http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack-dev