Re: administrivia (on avoiding injury)

[Greg Maxwell <gmaxwell@martin.fl.us>]

On Wed, 11 Apr 2001, 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. 
> 
                   
 HOST2--SITE2------ISPD
           \-ISPC   |
>                \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?

I've completed your graph.
I'll assume there is multihoming smarts in the hosts (perhaps as much or
				    more then SCTP, but not necessarily).

ISPA:  DEAD:A000::/52
ISPB:  DEAD:B000::/52
ISPC:  DEAD:C000::/52
ISPD:  DEAD:D000::/52

SITE: DEAD:A001::/64 DEAD:B001::/64
SITE2: DEAD:C001::/64 DEAD:D001::/64

Host:  DEAD:A001::B0B1 DEAD:B001::BOB1
Host2: DEAD:C001::0ED0 DEAD:D001::0ED0

Possible paths:

DEAD:A001::B0B1 DEAD:C001::0ED0
DEAD:B001::BOB1	DEAD:C001::0ED0
DEAD:A001::BOB1 DEAD:D001::0ED0
DEAD:B001::BOB1 DEAD:D001::0ED0

Initating host will get dest addresses via DNS, pick a pair, start
talking, if nothing happens, it will try another.

Once talking, all is happy until something fails...

Retransmit is tried against one of the other pairs (later ICMP errors
could be used to make the retry process smarter and faster) until
connectivity is regained.

The SCTP RFC covers a complete protocol implimentation that can do this.