Re: administrivia (on avoiding injury)

[Margaret Wasserman <mrw@windriver.com>]

Good questions!

In current IPv4 multi-homing solutions, the answer is that the
host does _not_ detect that the link is dead.  The routing
infrastructure detects that the link is dead and re-routes
traffic accordingly.  Right?

According to our current requirements, I think that there are
several options for IPv6 multi-homing:

         (1)  The host does detect that the link is dead and
                 intentionally moves to a new address on a "live"
                 link. (This would require stack changes, some 
                 sort of dead-link notification mechanism and/or
                 routing knowledge on the host).
         (2)  The host does not detect a dead link, but does
                 detect that the communication is not progressing
                 and initiates fail-over to alternate addresses
                 to keep current communications alive.
                 (Requires application, DNS and possibly stack
                 changes on the host).
         (3)  The host does not detect a dead link and the 
                 current communication fails.  However, the routing
                 system has deprecated the prefixes on the "broken"
                 ISP, so new connections will use working addresses.

It is my hope that we could have a solution that includes either
(1) or (2) for new IPv6 hosts AND (3) for current IPv6 hosts.

Margaret


At 10:47 PM 4/10/01 , Sean Doran wrote:
>Before you implement, please answer these three questions:
>
>| > Personally, I think it's reasonable to require a host change to switch to
>| > another prefix when there is link failure. 
>
>                 World----ISPZ
>                   |       |
>                 ISPA     ISPB
>                     \   /
>                     site
>                       |
>                     host
>
>How does host detect a link failure between ISPB and ISPZ?
>How does host detect a link failure between ISPZ and World?
>When host detects either of these failures, how does it adapt
>such that there is no lost connectivity to what can still be 
>reached through site's _still operating_ connection to ISPB?
>
>         Sean.