On Wed, Nov 11, 2009 at 3:19 PM, William Herrin
<bill at herrin.us> wrote:
Hi folks,
I couldn't attend but I had two questions about the FIB aggregation
draft (http://tools.ietf.org/html/draft-zhang-fibaggregation-02):
1. The level 3 and 4 algorithms disregard the implicit default, right?
Yes.
A destination that falls through the routing table without matching a
route is unreachable. You can think of this in terms as an implicit
route for 0.0.0.0/0 with a next hop of: local process, compose ICMP
unreachable. If I understood the paper, some or all of the address
space within 0.0.0.0/0 which would otherwise hit that implicit default
will be assigned an essentially random next hop, with the system
depending on the TTL to eventually drop the packet.
Or, the packets may be dropped if the next router does not assign any next hop to this "extra routable space", as different routers aggregate differently.
Have level 3 and level 4 been tested after adding an explicit default
that has to be honored to see what difference it makes in the numbers?
No, we didn't add any explicit default route. We can give it a try to see what numbers come out. But intuitively, adding a default route means taking away all extra route space, so the best aggregation can be achieved by the ORTC algorithm mentioned by John in an early email. It should be close to what we have for Level-2 aggregation, which is about 50% table size.
2. If I understand figure 6 correctly, the numbers are the percentage
of original size the FIB reduces to. So, 2008 level 4b, if you started
with a 1000 entry FIB, after aggregation the FIB still contains 363
entries. Did I read it correctly?
Yes.
Regards,
---
beichuan
Thanks,
Bill Herrin
--
William D. Herrin ................ herrin at dirtside.com bill at herrin.us
3005 Crane Dr. ...................... Web: <http://bill.herrin.us/>
Falls Church, VA 22042-3004
_______________________________________________
GROW mailing list
GROW at ietf.org
https://www.ietf.org/mailman/listinfo/grow