[Openstack-operators] Neutron HA

Jacob Godin jacobgodin at gmail.com
Thu Jul 18 12:25:46 UTC 2013


I guess my question would be how to get it to work with Pacemaker at the L2
level w/ OVS. There are scripts available for L3, DHCP, and Quantum Server,
but nothing for any of the L2 agents.


On Wed, Jul 17, 2013 at 4:46 PM, Daneyon Hansen (danehans) <
danehans at cisco.com> wrote:

>   Jacob,
>
>  The two options I'm aware of are:
>
>  1.  Active/Active using Quantum L2 only with Provider Networking
> Extensions. In this scenario, you leverage the physical data center network
> to provide L3 redundancy through mechanisms such as VRRP.
> 2.  Active/Passive using Pacemaker.
>
>  Regards,
> Daneyon Hansen
> Software Engineer
> Email: danehans at cisco.com
> Phone: 303-718-0400
> http://about.me/daneyon_hansen
>
>   From: Jacob Godin <jacobgodin at gmail.com>
> Date: Wednesday, July 17, 2013 1:10 PM
> To: Cisco Employee <danehans at cisco.com>
> Cc: "openstack-operators at lists.openstack.org" <
> openstack-operators at lists.openstack.org>
> Subject: Re: [Openstack-operators] Neutron HA
>
>   Thanks for the heads up. Would there be any solution that would take
> advantage of Linux HA or something?
>
>
>
> On Wed, Jul 17, 2013 at 3:09 PM, Daneyon Hansen (danehans) <
> danehans at cisco.com> wrote:
>
>>   Jacob,
>>
>>  At this time, not with the L3 Agent.  I suggest following this bug:
>>
>>  https://bugs.launchpad.net/neutron/+bug/1154622
>>
>>
>>  Regards,
>> Daneyon Hansen
>> Software Engineer
>> Email: danehans at cisco.com
>> Phone: 303-718-0400
>> http://about.me/daneyon_hansen
>>
>>   From: Jacob Godin <jacobgodin at gmail.com>
>> Date: Wednesday, July 17, 2013 11:37 AM
>> To: "openstack-operators at lists.openstack.org" <
>> openstack-operators at lists.openstack.org>
>> Subject: [Openstack-operators] Neutron HA
>>
>>   Hi all,
>>
>>  Has anyone successfully deployed a truly HA setup of Neutron (Quantum)?
>> I've been reading about the agent scheduling functionalities etc., but they
>> do not seem to provide true HA in a failure scenario. Particularly, this
>> article lays it out nicely:
>> http://www.mirantis.com/blog/a-new-agent-management-approach-in-quantum-for-openstack-grizzly/
>> .
>>
>>  I'm using OVS, and have a dedicated network node with my OVS, dhcp, and
>> L3 agents running on it. What I want to do, is deploy those services to my
>> "secondary" (or second active) networking node. In the case of a failure,
>> this node should pick up any networking tasks that the other node was
>> responsible for.
>>
>>
>>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.openstack.org/pipermail/openstack-operators/attachments/20130718/492ff838/attachment.html>


More information about the OpenStack-operators mailing list