yahoo-eng-team team mailing list archive

[Bolke de Bruin <bolke.de.bruin@xxxxxx>]

Public bug reported:

I've been struggling with this issue since the release of folsom, the
first install. I'm running havana now and it still turns up.

My network setups looks like this:
- GRE setup
- Directly connected interfaces to external

Each compute node has two interfaces:
- Mgmt network (GRE routing goes here)
- External network (no ip address assigned)

Example network:

192.168.248.0/24 (Internal)
192.168.248.1 router (internal interface)
172.17.11.71 router (external interface)
192.168.248.3 DHCP
192.168.248.4 VM

After a reboot of the network node I have to bring the admin state of
the internal interface of the route down and then up (it was active). If
i do this manually eg. "ip nets exec qrouter-XXX exec ifconfig tapxxxx
up/down it doesn't work. So something else gets changed.

Moreover, sometimes it is required to do the same for the DHCP interface
and even the VM interface before packets are being accepted/transmitted.

I checked logs, tcpdumps, upgraded openvswitch and kernels. I dont have
a clue any more really.

** Affects: neutron
     Importance: Undecided
         Status: New


** Tags: connectivity interface neutron qrouter reboot

-- 
You received this bug notification because you are a member of Yahoo!
Engineering Team, which is subscribed to neutron.
https://bugs.launchpad.net/bugs/1323079

Title:
  Some network ports "down" after reboot netnode

Status in OpenStack Neutron (virtual network service):
  New

Bug description:
  I've been struggling with this issue since the release of folsom, the
  first install. I'm running havana now and it still turns up.

  My network setups looks like this:
  - GRE setup
  - Directly connected interfaces to external

  Each compute node has two interfaces:
  - Mgmt network (GRE routing goes here)
  - External network (no ip address assigned)

  Example network:

  192.168.248.0/24 (Internal)
  192.168.248.1 router (internal interface)
  172.17.11.71 router (external interface)
  192.168.248.3 DHCP
  192.168.248.4 VM

  After a reboot of the network node I have to bring the admin state of
  the internal interface of the route down and then up (it was active).
  If i do this manually eg. "ip nets exec qrouter-XXX exec ifconfig
  tapxxxx up/down it doesn't work. So something else gets changed.

  Moreover, sometimes it is required to do the same for the DHCP
  interface and even the VM interface before packets are being
  accepted/transmitted.

  I checked logs, tcpdumps, upgraded openvswitch and kernels. I dont
  have a clue any more really.

To manage notifications about this bug go to:
https://bugs.launchpad.net/neutron/+bug/1323079/+subscriptions