Open Stack

Nihan,

I'm a bit of a newbie and your answer is rather cryptic.  What does that
mean "I have to add to guest (instance)"?  Add what to the guest?  How do I
do that?  (Guestt is running cirros)

Thank you


Jeff



On Fri, Aug 15, 2014 at 7:02 PM, Nhan Cao <nhanct92 at gmail.com> wrote:

> you must add to guest (instance)
>
>
> 2014-08-16 1:43 GMT+07:00 Jeff Silverman <jeff at sweetlabs.com>:
>
> I have been surfing the internet, and one of the ideas that comes to mind
>> is modifying the /etc/neutron/agent.ini file on the compute nodes.  In the
>> agent.ini file, there is a comment near the top that is almost helpful:
>>
>> # L3 requires that an interface driver be set. Choose the one that best
>> # matches your plugin.
>>
>> The only plug I know about is ml2.  I have no idea if that is right for
>> me or not.  And I have no idea to choose the interface drive that best
>> matches my plugin.
>>
>> Thank you
>>
>>
>> Jeff
>>
>>
>>
>>
>>
>>
>> On Fri, Aug 15, 2014 at 10:26 AM, Jeff Silverman <jeff at sweetlabs.com>
>> wrote:
>>
>>> By "defined a network space for your instances", does that mean going
>>> through the process as described in
>>> http://docs.openstack.org/icehouse/install-guide/install/yum/content/neutron-ml2-compute-node.html
>>> ?
>>>
>>> I got part way through that when I realized that the procedure was going
>>> to bridge packets through neutron.  That's not what I want.  I want the
>>> packets to go directly to the physical router.  For example, I have two
>>> tenants, with IP addresses 10.50.15.80/24 and 10.50.18.15.90/24.and the
>>> router is at 10.50.15.1.  There is a nice picture of what I am trying to do
>>> at
>>> http://docs.openstack.org/openstack-ops/content/network_troubleshooting.html#nova_network_traffic_in_cloud
>>> .  But if the hypervisor doesn't present a virtual device to the guests,
>>> then nothing else is going to happen.  The network troubleshooting guide
>>> http://docs.openstack.org/openstack-ops/content/network_troubleshooting.html#nova_network_traffic_in_cloud
>>> does not explain what to do if the virtual NIC is missing.
>>>
>>>
>>> Thank you
>>>
>>> Jeff
>>>
>>>
>>>
>>> On Fri, Aug 15, 2014 at 9:38 AM, Abel Lopez <alopgeek at gmail.com> wrote:
>>>
>>>> Curious if you’ve defined a network space for your instances. If you’re
>>>> using the traditional flat_network, this is known as the ‘fixed_address’
>>>> space.
>>>> If you’re using neutron, you would need to create a network and a
>>>> subnet (and router with gateway, etc). You’d then assign the instance to a
>>>> network at launch time.
>>>>
>>>>
>>>> On Aug 15, 2014, at 9:17 AM, Jeff Silverman <jeff at sweetlabs.com> wrote:
>>>>
>>>> <ip_a.png>
>>>> ​
>>>> For those of you that can't see pictures:
>>>> $ sudo ip a
>>>> 1: lo: <LOOPBACK,UP,LOWER_UP> mtu 16436 qdisc noqueue
>>>>     link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
>>>>     inet 127.0.0.1/8 scope host lo
>>>>     inet6 ::1/128 scope host
>>>>         valid_lft forever preferred_1ft forever
>>>>
>>>> I suspect that the issue is that the hypervisor is not presenting a
>>>> virtual ethernet card.
>>>>
>>>> Thank you
>>>>
>>>>
>>>> Jeff
>>>>
>>>>
>>>>
>>>> On Thu, Aug 14, 2014 at 6:57 PM, Nhan Cao <nhanct92 at gmail.com> wrote:
>>>>
>>>>> can you show output of command:
>>>>> ip a
>>>>>
>>>>>
>>>>>
>>>>>
>>>>> 2014-08-15 7:41 GMT+07:00 Jeff Silverman <jeff at sweetlabs.com>:
>>>>>
>>>>>> People,
>>>>>>
>>>>>> I have brought up an instance, and I can connect to it using my
>>>>>> browser!  I am so pleased.
>>>>>>
>>>>>> However, my instance doesn't have an ethernet device, only a loopback
>>>>>> device.   My management wants me to use a provider network, which I
>>>>>> understand to mean that my instances will have IP addresses in the same
>>>>>> space as the controller, block storage, and compute node administrative
>>>>>> addresses.  However, I think that discussing addressing is premature until
>>>>>> I have a working virtual ethernet card.
>>>>>>
>>>>>> I am reading through
>>>>>> http://docs.openstack.org/icehouse/install-guide/install/yum/content/neutron-ml2-compute-node.html
>>>>>> and I think that the ML2 plugin is what I need.  However, I think I do not
>>>>>> want a network type of GRE, because that encapsulates the packets and I
>>>>>> don't have anything to un-encapsulate them.
>>>>>>
>>>>>> Thank you
>>>>>>
>>>>>>
>>>>>> Jeff
>>>>>>
>>>>>>
>>>>>> --
>>>>>> *Jeff Silverman*
>>>>>> Systems Engineer
>>>>>> (253) 459-2318 (c)
>>>>>>
>>>>>>
>>>>>> _______________________________________________
>>>>>> OpenStack-operators mailing list
>>>>>> OpenStack-operators at lists.openstack.org
>>>>>>
>>>>>> http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack-operators
>>>>>>
>>>>>>
>>>>>
>>>>
>>>>
>>>> --
>>>> *Jeff Silverman*
>>>> Systems Engineer
>>>> (253) 459-2318 (c)
>>>>
>>>>  _______________________________________________
>>>> OpenStack-operators mailing list
>>>> OpenStack-operators at lists.openstack.org
>>>> http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack-operators
>>>>
>>>>
>>>>
>>>
>>>
>>> --
>>> *Jeff Silverman*
>>> Systems Engineer
>>> (253) 459-2318 (c)
>>>
>>>
>>
>>
>> --
>> *Jeff Silverman*
>> Systems Engineer
>> (253) 459-2318 (c)
>>
>>
>


-- 
*Jeff Silverman*
Systems Engineer
(253) 459-2318 (c)
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.openstack.org/pipermail/openstack-operators/attachments/20140818/b573e0e5/attachment.html>