WAY TOO BAD route flapping on 6bone

itojun@itojun.org itojun@itojun.org
Fri, 03 Mar 2000 19:50:53 +0900


	The issue seem to be a combination of EBGP flapping (1 update
	per second), and local IBGP flapping due to a bug in route reflector
	(more than 3 incoming update per second, result in lots more flood
	of updates).  We (WIDE) have large inside topology and it contributed
	to the particular flaps.
	It looks the prefixes listed are not directly related to severe flaps
	in IBGP, it could have happened for any of the prefixes we receive.
	We are trying to look carefully after updating the code we use.

	Sorry for confusions.

itojun