>> The Microsoft implementation limits to two errors per second, just for
>> this reason.
>
> That is only good enough if you assume the tracerouter is over 500ms away, 
> because traceroute by default sends out three probes..
>
> I've commented on the new ICMP draft that this issue should be taken into
> consideration, and apparently it will, as the draft didn't progress at
> IETF51.
>
> B.t.w. Linux uses a token bucket mechanism here and is not affected.
>
> B.t.w.2. Cisco implementation is(/was) also broken, another issue I
> mentioned here and on [EMAIL PROTECTED], they promised they'd be working on
> it.

the latest releases use an adjustable token bucket algorithm.

/ot
--------------------------------------------------------------------
IETF IPng Working Group Mailing List
IPng Home Page:                      http://playground.sun.com/ipng
FTP archive:                      ftp://playground.sun.com/pub/ipng
Direct all administrative requests to [EMAIL PROTECTED]
--------------------------------------------------------------------

Reply via email to