[Mtools-report] First list of targets

[ken lindahl]

John Haskins wrote:
> I ran Jeremy's list through awk and put up a kick-me smokeping instance 
> at: 
> https://dogbert.commserv.ucsb.edu/cgi-bin/smokeping.cgi?target=CENIC-mtools-pilot

thanks for doing this, John.

there's something that strikes me as odd though: for all of the
sites that i've checked [*], the median RTT is graphed around a
baseline ("avg") that is exactly 10.0ms, or 20.0ms, or 30.0ms.
i.e. there are none where the "avg" median RTT is some value that
is not an even multiple of 10.0.

why is this? does smokeping have a "startup rule" underwhich
for some time after a new target is added, it uses an artificial
value for the baseline; and later, after enough measurements
are collected, it starts using the measured value of the average?

[*] i checked a fair number of the sites, but not all of them.
it's entirely possible that i missed some sites that disprove
my conjecture. i agree that the list is too long and unwieldy,
i agree with someone's suggestion earlier of arranging them
hierarchically by region/POP).

thanks,
ken