[Openstack] LBaaS solution for eleminate SPOF is required

Eugene Nikanorov enikanorov at mirantis.com
Tue Apr 1 09:45:51 UTC 2014


Hi Heiko,

That kind of failover mechanism was on our roadmap:
https://review.openstack.org/#/c/59743/
However right now the work is deferred as we're discussing general design
of HA feature.
I can't guarantee that will be something included in Juno, but there's
strong demand for this feature so it has high priority on our list.

Thanks,
Eugene.


On Tue, Apr 1, 2014 at 1:21 PM, Édouard Thuleau <thuleau at gmail.com> wrote:

> Just point you a blueprint [1] develop during the I release (but deferring
> to Juno) to bring L3 HA.
> It use VRRP mechanism with keepalived that you could reuse for LBaaS.
> It also uses conntrackd to synchronize connections between routers (master
> and slave). I don't know how to configure LB synchronization.
>
> [1] https://blueprints.launchpad.net/neutron/+spec/l3-high-availability
>
> Édouard.
>
>
> On Tue, Apr 1, 2014 at 10:52 AM, Heiko Krämer <hkraemer at anynines.com>wrote:
>
>>
>> -----BEGIN PGP SIGNED MESSAGE-----
>> Hash: SHA1
>>
>> Hi guys,
>>
>> i'm looking since days for a solution to solve the SPOF (single point of
>> failure) of the lbaas. A easy solution with a config option is not
>> available yet.
>>
>> The next option is to install/configure some not OpenStack related tools
>> like pacemaker. I realise my L3-Agent faiover with an self writen RA for
>> pacemaker to switch routers on broken l3 agent to the running on with
>> the neutron API.
>> The same can be workin with LbaaS but there's a problem again:
>>
>>  - No API function to switch a pool from agent A to agent b
>>  - configurations of HAProxy replicate => this can be solvd with drbd or
>> some shared network storage like NFS
>>  - Devices are not the same => L3 switch router from agent A (broken) to
>> agent B the devices names are not the same
>>  -
>>
>> Is there any solution to solve this SPOF ? I mean a LB should not only
>> scale/balancing your cluster/application the second task is the failover
>> of broken members.
>>
>>
>>
>>
>> Thx and cheers
>> Heiko
>>
>> - --
>> anynines.com
>> -----BEGIN PGP SIGNATURE-----
>> Version: GnuPG v1.4.14 (GNU/Linux)
>> Comment: Using GnuPG with Thunderbird - http://www.enigmail.net/
>>
>> iQEcBAEBAgAGBQJTOn4wAAoJELxFogM4ixOFUxoIAIb5WYnRNimu6C67SyueCs7w
>> BkBee1atnIyhBPv7rz+5dHG+JQbSPHzmh+yK6kXgXGrDkxqLRDcd9o+Exv9oRUe0
>> 4TU79YCgN2wqIHP5mhioHyd5f0V4MYjoLHDQRYgFSSXf0c4uSGdIvWRxFsSY8H9q
>> 6nGF35mTOUd/Owu2EQ82SGWL2oOauG+G9QC79CNjIex4ECdtUZfCuxGV2ZVJjz+L
>> EhsPWWy+D+qvtE+9+rRCpo8Xv4PNY+8Hj3S+MPppzD3SUmchlo6R9DXEbBQqE7g3
>> VvjzyfclT+hjz3KbVdR8P96uMkcYQNDiJqPz1EUevcZRfnjPSJu76VvI2bOkcBo=
>> =nojq
>> -----END PGP SIGNATURE-----
>>
>>
>> _______________________________________________
>> 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
>>
>
>
> _______________________________________________
> 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
>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.openstack.org/pipermail/openstack/attachments/20140401/7688816f/attachment.html>


More information about the Openstack mailing list