[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: [RRG] Which Side to Control Ingress Link Selection?

Scott -

Apologies for the delay in responding.  You made very convincing
arguments against assuming complete knowledge of other edge net-
work's egress indirection routers.

Coming back to the initial problem of whether and how to ensure
that packets from a sending edge network take the ingress link
selected by the receiving edge network, independent of which egress
indirection router of the sending edge network these packets go
through, and re-citing the two solution options I stated earlier...


(a) Synchronization between indirection routers in the sending edge
   network.  The recipient edge network must then send its
   reachability information to only one indirection router of the
   sending edge network, and this would likely be the indirection
   router that handles packets in the other direction.

(b) Sending reachability information to all indirection routers of a
   sending edge network.


...the result of this discussion is then that solution option (a) is
preferable because it does not require the receiving edge network to
know about all egress indirection routers of the sending edge network.

The synchronization at the sending edge network could happen via
multicast, where each egress indirection router joins the multicast
group.

The only issue I see with this solution is that there is no convincing
incentives model: Why would the sending edge network pursue the synchronization? There is a small incentive if we consider fail-over 
indications from the receiving edge network (e.g., when one of the re-
ceiving edge network's providers goes down).  The sending edge network
then better informs all egress indirection routers quickly about
reachability updates from the receiving edge network.  But if the re-
ceiving edge network is sending reachability updates just for local
optimization purposes (e.g., load-balancing on ingress links), then
the sending edge network has no incentive to spread that information
to all of its egress indirection routers.

Finally, for completeness, there is a 3rd solution option, which
I haven't mentioned before:

(c)  Assume a pull-based mapping resolution system where mappings
     are retrieved from the edge network that holds the mapping (e.g.,
     TRRP or DNS Map).  Then an egress indirection router that requires
     fresh reachability information for a receiving edge network could
     perform mapping resolution to determine the receiving edge net-
     work's preferred ingress link.

Note that this solution option touches the recent discussion on desired
properties of mapping resolution systems...

- Christian



--
to unsubscribe send a message to rrg-request@psg.com with the
word 'unsubscribe' in a single line as the message text body.
archive: <http://psg.com/lists/rrg/> & ftp://psg.com/pub/lists/rrg