[one-users] Oned failover mechanism and multi-cluster deployment ?
timm at fnal.gov
Fri Nov 12 10:57:40 PST 2010
I have seen the new feature about logical clusters
(the onecluster command) but it wasn't clear what
capacities it really brings. Is there any benefit to
defining more than one cluster with the onecluster command,
if so, what?
On Fri, 12 Nov 2010, Tino Vazquez wrote:
> Hi Ivan,
> We are currently working on different levels of fault tolerance. We
> want to have (in the short term) a best practices document to achieve
> fault tolerance with OpenNebula.
> About a multi-cluster deployment architecture, there is currently two
> ways of achieving this:
> * Logic clusters: this is a new feature introduced in the 2.0 version.
> See  for details.
> * Federation can be achieved using plugins, where for instance the
> Amazon EC2 driver and server is used to federate different clusters,
> each one of them managed by one OpenNebula instance.
> Hope it helps,
> Constantino Vázquez Blanco | dsa-research.org/tinova
> Virtualization Technology Engineer / Researcher
> OpenNebula Toolkit | opennebula.org
> On Mon, Oct 25, 2010 at 2:17 PM, Ivan Frain <ivan.frain at gmail.com> wrote:
>> Hi everybody,
>> I would like to know if there is a failover mechanism for oned ? I'am sure
>> that it is possible to manage this failover using MySQL as database backend
>> on a remote server but I would like to have more details: should we use
>> active/standby mechanism with heartbeat, is there any guidelines to do this
>> Moreover, in the case of oned failure, what are the guarantees about its
>> state and the recovery?
>> I didn't find any documentation in the OpenNebula web site this is why I
>> post my question here.
>> I have another question concerning a multi cluster management. I understand
>> the deployment of opennebula in one cluster with one frontend node on which
>> 'oned' is installed but what could be a deployment with several physical
>> clusters cooperating each one running opennebula toolkit? Can we put in
>> place a kind of hierarchy like Eucalyptus: cloud controller, cluster
>> controller and so on ?
>> Thank you in advance.
>> Best regards,
>> Ivan Frain
>> Users mailing list
>> Users at lists.opennebula.org
> Users mailing list
> Users at lists.opennebula.org
Steven C. Timm, Ph.D (630) 840-8525
timm at fnal.gov http://home.fnal.gov/~timm/
Fermilab Computing Division, Scientific Computing Facilities,
Grid Facilities Department, FermiGrid Services Group, Assistant Group Leader.
More information about the Users