I have noticed that the site http://www.internettrafficreport.com/30day.htm
shows a drastic reduction in the global packet loss rate since 28 October
2005 (the reported rate fell from a year long 16-20% loss rate down to a new
and pleasant ~1% loss rate).

Does anyone know what caused this reduction in reported packet loss? Is it a
real reduction or just a problem with internettrafficreport ?

Years ago I did a lot of work on a IP link that had widely varying packet
loss. I noticed that many of the high level protocols (TELNET, FTP, etc
would fail when the packet loss exceeded around 12%, due to repeated
timeouts waiting for lost packets. It is a delight to see the reported
present loss of ~1%, but it seems too good to believe.

What's up?

Jerry