Pchar is a tool to characterize the bandwidth, latency, and loss of links along an end-to-end path through the Internet. It is based on the algorithms of the pathchar utility written by Van Jacobson, formerly of Lawrence Berkeley Laboratories.

Pchar measures the characteristics of the network path between two Internet hosts, on an IPv4 or IPv6 network. The program measures network throughput and round trip time by sending UDP packets of varying size into the network and waiting for ICMP messages in response. It modulates the IPv4 time to live (TTL) field or the IPv6 hop limit field to get measurements at different distances along a path.

Pchar presents the following details for each hop in the trip:

• the number of partial or lost datagrams and percentage of probe packets that were lost during the probes for that hop
• the estimated round trip time from the probing host through the current hop
• estimates of the round trip time and bandwidth for the current hop
• estimate of the average queuing along the path, up to and including the current hop

After the last hop (usually the target host), pchar prints statistics in the entire path, including the path length and path pipe (the latter is an estimate of the delay bandwidth product of the path).

In the other/second mode of operation called “trout” (short for “tiny traceroute”) Pchar sends packets of random sizes (one packet per hop diameter) along the path to a destination. This mode is extremely fast but no attempt at estimating link properties is made.

Example implementation in 6NET:

In PSNC, Pchar is running on a host called plum.man.poznan.pl. It can be accessed directly using the following link: http://plum.man.poznan.pl/cgi-bin/pchar.cgi . The user-friendly web interface for the Pchar tool is implemented as a Perl script written in PSNC. Their implementation is IPv4 and IPv6 enabled and publicly available through the links mentioned above. Only basic options for this tool are used to simplify the user interface. During the testing phase and every day use PSNC have noticed that Pchar can return quite accurate results for the nearest neighbours, but for far away hosts it can give incorrect results. Generally, Pchar can estimate the correct bandwidth for the near neighbours before it reaches the bottleneck of this link. Unfortunately, it can then propagate any limitations and approximations caused by the link’s bottleneck giving inexact and unreliable results.