On Fri, 25 Jul 2014, valdis.kletni...@vt.edu wrote:
On Fri, 25 Jul 2014 14:20:53 -0700, David Lang said:
cost of bandwidth for this is just something to get someone to pay for (ideally
someone with tons of bandwidth already who won't notice this sort of test, even
if there are a few going on at once.)
Ask U of Wisconsin how that worked out for them when Netgear shipped some
new boxes....
http://en.wikipedia.org/wiki/NTP_server_misuse_and_abuse#NETGEAR_and_the_University_of_Wisconsin.E2.80.93Madison
It's one thing to come up with a solution that works for the 300 (total wild
guess) people
on this list. But to be useful for field deployable, it really needs to be able
to handle a "every Comcast customer in a major metro area", because we *do*
want this stuff to work well when this makes it into the CPE that Comcast
gives every new customer, right? :)
the key thing here is to stagger the work.
Don't have every device do the test immediatly at startup. As you noted, that
will collapse as they all start up at once.
Have them operate from their saved stats and test at (largish) random offsets.
ideally, have this test be against an ANYCAST address, so that the ISP can run a
copy internally without having to reconfigure the clients.
David Lang
_______________________________________________
Cerowrt-devel mailing list
Cerowrt-devel@lists.bufferbloat.net
https://lists.bufferbloat.net/listinfo/cerowrt-devel
