Hi,
I had the same problem.
My problem was that the openvswitch plugin's config file was missing on
the controller node, and because of this, the horizon used "local"
network not "gre" (the same problem can be the problem with VLAN setups)
you can chek it with neutron net-show.
I hope it helps
Chen ,
I think ur packets is getting dropped. U can verify this by just
checking the packets coming from compute node hitting which of the rules in
br-int.
On Wed, Feb 12, 2014 at 10:56 AM, Robert Collins
wrote:
> Check the outbound ofctl rules on your hypervisor nodes. If they
> aren't tag
Check the outbound ofctl rules on your hypervisor nodes. If they
aren't tagging traffic properly, it won't be processed by the incoming
gre rules and you'll see the symptoms you have.
On 8 February 2014 05:41, Chris Baker wrote:
> Hi guys,
>
> My havana installation has 3 nodes:
> control node, r
=NORMAL
Thanks.
-chen
From: Vikash Kumar [mailto:vikash.ku...@oneconvergence.com]
Sent: Wednesday, February 12, 2014 12:41 PM
To: Chris Baker
Cc: openstack
Subject: Re: [Openstack] dhcp request can't reach br-int of network node
U can try few things quickly.
a) Connectivity between ur compute
U can try few things quickly.
a) Connectivity between ur compute node and n/w node. Check the interfaces.
b) Check the route entry.
Ur compute node can ping to n/w node. U can also chk the vlan tag settings
on ovs bridge.
On Fri, Feb 7, 2014 at 10:11 PM, Chris Baker wrote:
> Hi guys,
>
> My h
It's most likely you need to look inside the namespaces to see the
traffic on br-int, are you familiar with 'ip netns exec'? I just
added a bit about this to the operators guide
http://docs.openstack.org/trunk/openstack-ops/content/network_troubleshooting.html#dealing_with_netns
there's a bit more
