Open Stack

It's sent via http. Look for the nova_url parameter.[1]

1.
https://github.com/openstack/neutron/blob/33b83554442b7c0d862a2adb0b2a8feab1185af9/etc/neutron.conf#L267


On Thu, Feb 5, 2015 at 11:10 PM, Akilesh K <akilesh1597 at gmail.com> wrote:

> Great this is what I was looking for and I was suspecting the same. After
> instance launch everything is smooth , dhcp, routing , floatingip. So I
> believe nova is just missing the notification from neutron. How does this
> notification go, through rabbitmq?
>
> Thank you,
> Ageeleshwar K
>
> On Fri, Feb 6, 2015 at 12:36 PM, Kevin Benton <blak111 at gmail.com> wrote:
>
>> Nova sends a port update to Neutron after the VM has been scheduled so
>> Neutron can know the location of the port it needs to setup in OVS. After
>> Neutron finishes this setup, it notifies Nova that it has finished setting
>> up the port. At this point, Nova resumes the boot of the VM since network
>> connectivity is now supposed to be present.
>>
>> vif_plugging_timeout is how long Nova will wait to hear back from Neutron
>> that the port is active.
>> vif_plugging_is_fatal determines what Nova does when that timer expires.
>> If it's true, it will error out as you observed. If it's false, it will
>> assume the port is ready and continue booting even though it has not heard
>> from Neutron.
>>
>> Your intuition is right that you are bypassing the problem, which is the
>> missing notifications from Neutron to Nova. Have you confirmed that the
>> Neutron port is actually active and that you have connectivity?
>>
>> If so, there is probably just a misconfiguration in neutron.conf for the
>> nova notification settings. Check the URL/region/tenant
>> info/username/password for all of the settings that start with 'nova_' in
>> neutron.conf.
>>
>> If you don't have connectivity, it means that Neutron actually couldn't
>> bring the port online, in which case it was correct not to notify Nova. In
>> this case, you will need to examine your Neutron server logs (usually
>> /var/log/neutron/server.log) to get more hints about why it's failing to
>> bring up the port.
>>
>> Cheers,
>> Kevin Benton
>>
>> On Thu, Feb 5, 2015 at 10:06 PM, Akilesh K <akilesh1597 at gmail.com> wrote:
>>
>>> Hi,
>>>
>>> I have a working openstack setup. But I have some questions on how I got
>>> it working.
>>>
>>> Initially my instances failed to launch.
>>>
>>> I have updated the correct values for neutron authentication in
>>> nova.conf and correct nova authentication in neutron.conf and notify_nova
>>> on port change is true but still instances failed.
>>>
>>> One thing I did note was that It took an awfully long time to fail and
>>> by the time it reported failure neutron had created a port for the
>>> instance, the vif driver had created a tap interface, the kvm had already
>>> started but still nova reported failure.
>>>
>>> When I set vif_plugging_is_fatal to false and vif_plugging_timeout to 10
>>> I could finally have my Instances up and in about 20 seconds.
>>>
>>> Someone please explain the
>>> 1, significance of these two parameters
>>>
>>> 2. what had been going on in my openstack before I had them configured
>>> correctly
>>>
>>> 3. if even the configuration I had done is correct or am I just
>>> bypassing some bigger problem.
>>>
>>> Thank you,
>>> Ageeleshwar K
>>>
>>> _______________________________________________
>>> Mailing list:
>>> http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack
>>> Post to     : openstack at lists.openstack.org
>>> Unsubscribe :
>>> http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack
>>>
>>>
>>
>>
>> --
>> Kevin Benton
>>
>
>


-- 
Kevin Benton
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.openstack.org/pipermail/openstack/attachments/20150205/d790b366/attachment.html>