MAAS server had two interfaces eth0 external net and eth1 internal (for pxe boot). Somehow I am seeing a virbr0 under interfaces and an IP assigned to it. Not sure where that interface and its IP came from. Is it possible it was created after juju install? Or while trying to commission node in MAAS. MAAS and JUJU are on same machine.

IP for virbr0 is similar for another post on this similar topic.

virbr0    Link encap:Ethernet  HWaddr d6:79:21:a9:19:6e
          inet addr:  Bcast:  Mask:
          UP BROADCAST MULTICAST  MTU:1500  Metric:1
          RX packets:0 errors:0 dropped:0 overruns:0 frame:0
          TX packets:0 errors:0 dropped:0 overruns:0 carrier:0
          collisions:0 txqueuelen:0
          RX bytes:0 (0.0 B)  TX bytes:0 (0.0 B)

IIRC on Ubuntu virbr0 with is automatically created when you install qemu packages. Run:

dpkg -l | grep -i qemu

to see is that the case.

  • Yes...i do see....can this interface be removed and will not impact MAAS...juju or landscape etc.....Thank you – aavam Apr 28 '16 at 15:30
  • Well, neither Juju, nor Landscape or MAAS requires Qemu (or libvirt). Why is it there in the first place? Were you planning to create virtual machines on MAAS server? If yes, then you should not remove this interface (bridge actually), because that will affect you hypervisor setup. – zalmarge Apr 28 '16 at 15:52
  • Cool....no plan to create VM in MAAS server..Although MAAS itself running on a Ubuntu 14.04.4 VM with eth0 --> needs proxy for internet or outside corp and eth1 --> 10.x.x.x interfaces......I have couple of blades that were pxe boot and ready in MAAS......Next step is to juju bootstrap and it is failinmg...juju bootstrap WARNING ignoring envs.yaml: using bootstrap config in file "/ho..nts/maas.jenv" ERROR cannot determine if envnt is already bootstrapped.: could not access file '032f91db-afda-4345-85d3-73bb8f99564c-provider-state': gomaasapi: got error back from server: 504 Gateway Timeout – aavam Apr 28 '16 at 17:05

As pointed out in above answer, virbr0 is a by-product of installing qemu which is useful to enable the MAAS controller to control libvirt-hosted "BNCs" (ie: KVM-hosted VMs). In my experience it's safe to get rid of virbr0; this can be achieved as follows:

virsh net-destroy default

(run as a user who's a member of the libvirtd group or sudo if necessary).

A side note: in my experience, the presence of virbr0 and associated subnet/fabric on the MAAS controller did cause some minor issues; for instance, for whatever reason provisioned nodes refer to the MAAS controller's IP on virbr0 in their ntp.conf (even though they don't have any NIC on that subnet); this may be a bug? (MAAS Version 2.1.3+bzr5573-0ubuntu1)

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.