>Hate to follow up on my own message, but here goes:
>
>After sending, I ran to google to see if I could find out what this
>problem was
>about.
>
>It seems that JUNOS had a dampening bug at one point where it would doubly
>penalize a flap, once at withdraw time, and again at re-advertise time.
>
>The link I turned up that describes the issues is at:
>http://www.cctec.com/maillists/nanog/current/msg00140.html
>
>This was from back in December, so I would imagine it's been corrected by now.
Even more than the NANOG list, the CISCO-NSP and JUNIPER-NSP mailing
lists discuss bugs. There is a difference between implementations
having bugs -- they all will -- and fundamental design problems. The
description you first described sounded like the Cisco engineer was
describing a fundamental design problem.
Quite frankly, one of the things I like about working in the carrier
space is that there's far less hype and FUD than in the enterprise
market. If I started touting the benefits of a "single-vendor,
end-to-end solution" to UUnet/MCI/MFS/Other members of the Worldcom
collective, Qwest, Level 3, Sprint, BBN/GTE/Verizon, etc., they'd
look at me as if I was out of my mind. (Not sure about that, some of
my technical ideas also get such a reaction at first).
The draft I'm updating on BGP convergence will be coauthored by me at
Nortel, Alvaro Retana at Cisco, Sue Hares and Padma Krishnaswami at
NextHop (the former GateD/Merit). We are actively trying to get a
coauthor from Juniper, and we've gotten reviews from Alcatel. The
fact that all of us want some common standards to be compared on is a
far cry from some of the salesy stuff that floats around in other
markets.
_________________________________
FAQ, list archives, and subscription info: http://www.groupstudy.com/list/cisco.html
Report misconduct and Nondisclosure violations to [EMAIL PROTECTED]
