NLANR is the largest project, which is publicly available for use by network
researchers, engineers, system administrators, and students. [10].
The goal of the NLANR Measurement and Network Analysis group is to study the
operation of the high performance connection (HPC) networks, measuring the flow
of traffic and characterizing the behaviour of these networks and improving the
design by solving the problem diagnosis to deliver maximum end-to-end
performance to users.
3.2 NLANR/AMP
NLANR/AMP is collecting IPv6 performance data in a mesh of eleven active
monitors [5]. There are nine monitors in US and two are placed internationally.
Figure 3 shows the location of AMP IPv6 monitors in US. From left to right:
University of Oregon, San Diego Supercomputer Center/UCSD, University of Utah,
University of Missouri at Columbia, University of Wisconsin, Michigan
Technological University, Georgia Institute of Technology, NYSERnet, Columbia
University.
Engineers perform measurements with the aid of the AMP IPv6 project. They have
been collecting IPv6 performance data since October 2002. They have found that
the general characteristics of IPv6 paths compared with IPv4 paths is that they
incur a large base delay, a jitter, and greater loss. They believe this is an
artifact of tunnel paths that follow less-than optimal paths, underpowered
tunnel entry and exit points, and an inefficient forwarding path inside some
routers that route IPv6 natively [5].
As they replaced IPv6-in-IPv4 with native IPv6 paths, they found that the IPv6
forwarding capability of routers improved.
They collect path and delay information using the IPv4 and IPv6 versions of ping
and traceroute, and compare the data on a path-by-path basis. The data and
related graphical representations for amp-Columbia on 28 May 2004 are shown in
Table 1.
By changing the IPv6 tunnel server from a Cisco GSR router located in Kansas
City to a Juniper T640 router in Indianapolis, they found that the new path was
topologically much closer to the path followed by IPv4 packets. This results in
a substantially lower base round trip time (RTT) delay, and less jitter as shown
in Figure 6 (the red colour represents IPv4 delay and the green colour
represents IPv6 delay). Figures 4 and 5 show the path change between amp-gatech
and amp-wisc, and which components of the path changed over the course of the
day (colours represent load levels of IPv6). As shown in Figure 4 the forward
path changes markedly after the tunnel change. The reverse path (see Figure 5)
is one hop shorter after the topology change, which is the result of two hops
being replaced with one new IPv6 hop [5].
Figure 4: The forward path, [http://amp.nlanr.net//IPv6/]
