hi all:
I have set up quantum+floodlight, there are a compute node and a
controller, so I create a VM in the compute node, but the VM(100.0.0.4) can
not ping its gateway(100.0.0.1) in the controller node.
When the VM send a ARP request to OVS of the compute node, a packet_in
request is sent to th
Hello,
I am getting confused in the architecture of ovs, the
netdev-provider is for hardware specific interfaces like eth0, and after
adding interface when traffic start it will go through datapath and
dpif-provider to userspace this means whenever packet comes to device dose
it go through
On May 6, 2013, at 4:09 PM, Luiz Ozaki wrote:
> On 5/6/13 7:21 PM, Ben Pfaff wrote:
>> On Mon, May 06, 2013 at 05:41:53PM -0300, Luiz Ozaki wrote:
>>> Production works fine BUT at the LAB the IN_PORT that I'm passing is
>>> 3 and it's matching with the IN_PORT 174 from my OF rule.
>> The in_port
On 5/6/13 7:21 PM, Ben Pfaff wrote:
On Mon, May 06, 2013 at 05:41:53PM -0300, Luiz Ozaki wrote:
Production works fine BUT at the LAB the IN_PORT that I'm passing is
3 and it's matching with the IN_PORT 174 from my OF rule.
The in_port value in this form of ofproto/trace is a kernel datapath
por
On Sat, May 04, 2013 at 08:02:15PM +0530, Sunayana Iyer wrote:
> Thanks for your help reid..
>
> yes now i am able to bring the ovsdbmonitor gui up, but it never shows
> anything, here is the error i am getting in logs.
>
> unix:/usr/local/var/run/openvswitch/db.sock '["Open_vSwitch",
> {"op":"s
On Tue, May 07, 2013 at 08:13:05AM +1000, Matthew Brown wrote:
> I'm trying to ping a VM (vnet0) attached to ovs-switch (br-int). My
> physical network is also attached to the ova-switch via eth1. I'm
> using POX as my controller running at 0.0.0.0:6633.
I guess you must have been the first person
Hello,
I am trying to understand how openstack created VM with a interface gets
added as a port into ovs-switch.
Anyone know how quantum server passes this information to the switch?
--
pals
___
discuss mailing list
discuss@openvswitch.org
http://o
On Mon, May 06, 2013 at 05:41:53PM -0300, Luiz Ozaki wrote:
> Production works fine BUT at the LAB the IN_PORT that I'm passing is
> 3 and it's matching with the IN_PORT 174 from my OF rule.
The in_port value in this form of ofproto/trace is a kernel datapath
port number, not an OpenFlow one. The
Hi all,
I'm trying to ping a VM (vnet0) attached to ovs-switch (br-int). My physical
network is also attached to the ova-switch via eth1. I'm using POX as my
controller running at 0.0.0.0:6633.
It seems that the VM is receiving the ARP broadcast and replying but that reply
doesn't get anywhere
Hello,
I'm making some changes on my controller so I'm getting some DP flows
from production and passing into my new controller/OVS (Which is awesome
by the way).
*Production:*
ovs-appctl ofproto/trace xapi5
"in_port(3),eth(src=56:d0:ec:e5:24:96,dst=02:e0:52:2a:ef:24),eth_type(0x0800),ipv4(
On Sun, May 5, 2013 at 6:36 PM, ZhengLingyun wrote:
> When I came across this problem, google it and found there is
>
> http://openvswitch.org/pipermail/discuss/2012-March/006603.html
>
> releated to it.
>
> As hamsiery said:
>
> No connectiving between VMs connected to differnet open vswitch brid
It would be non-trivial. If you really want to use 1.9.0, you could look at
porting Ben's original VXLAN patch:
http://openvswitch.org/pipermail/dev/2011-October/012051.html
It's not something we would include in 1.9.x maintenance releases or be able to
support, though.
--Justin
On
12 matches
Mail list logo
