Re: New charter item: HA assignment

Hi Vijay,

2006/7/18, Vijay Devarapalli <vijay.devarapalli@xxxxxxxxxxxxx>:
> Jean-Michel Combes wrote:
> > Hi,
> >
> > as promised during the mip6 session, here is a point which, IMO, is
> > important from a MSP's point of view:
> > Actually, the bootstrapping solution, in the split scenario, only
> > propose a HA discovery mechanism (i.e. the MN receives a list of HA,
> > from DNS, and chooses one of them). The MSP has no easy solution to
> > enforce one MN to have a specific HA.
>
> do you want an entity in the MSP to assign the HA
> to the MN?

I must admit I have not a solution in mind for the moment :(

> we can come up with some hacks to the HA discovery
> mechanism to force the MN to pick a particular HA,
> but I would rather avoid it. it would be better if
> you tell us what the scenario and which are the
> entities involved. then we can design a solution
> for the specific scenario.

A MSP wants to assign a specific HA to a specific MN because:
- the HAs are owned by the MSP (the main reason!!! :)
- the MN uses a specific mobility mechanism (e.g. MIPv6, HMIPv6, NEMO, Monami6)
- the operator wants to loadbalance the charge on the different HAs it
owns during the bootstrapping (e.g to avoid DoS attacks)
- the operator offers different types of QoS on the HAs


> another option is to use HA Switch mechanism described
> in http://www.ietf.org/internet-drafts/draft-ietf-mip6-ha-switch-00.txt

That could be a solution but I must admit I didn't read it recently.

Best regards.

JMC.

> Vijay