The number of hops on the route to this site ranged from 18 to 22. Mostly, packets took 18 hops to reach its destination port, but on 02/26/98 at 4:45 packets took 22 hops to reach its destination host. It is interesting to note that this run did not show the highest value for Max RTT. Loss rates were very high for this site. The loss rates varied from 0.766% to 40.08%. The range of loss percentage for the other sites is given below:

Bethesda: 0.05% to 2.66%
Canada: 0.01% to 1.35%
Chicago: 0% to 2.33%
Mexico: 0.075% to 18.02%
Taiwan: 1.26% to 15.22%

The average RTT value for this site ranged from 0.232 sec to 0.923 sec. This is very high when compared to the range of average RTT value for the other sites:

Bethesda: 0.015 to 0.025 sec
Canada: 0.039 to 0.089 sec
Chicago: 0.121 to 0.198 sec
Mexico: 0.122 to 0.411 sec
Taiwan: 0.309 to 0.432 sec

There is a slight variation in the minimum RTT value, ranging from 0.206 sec to 0.212 sec. The maximum RTT values are 9-15 times the minimum RTT value. They ranged from 1.916 sec to 3.255 sec. The number of reordered packets in this experiment ranged from 4 to 59. This was the only site that had duplicate packets. The number of duplicate packets ranged from 1 to 2. On 02/25/98 at 19:45, we observed that the roundtrip time periodically increased to approximately 0.50 sec at about 80 sec (2000 packet) intervals. For the same run, we observed that there was a periodic loss at about 30 sec (750 packet) interval. Note that the access to this site was via the GARR Network (a government network for universities and research institutes in Italy) and its bandwidth is 64kbps. Unfortunately, this network was, at that time, under heavy reconstruction and not very reliable.

[Bethesda] [Canada] [Chicago] [Mexico] [Italy] [Taiwan]

[Back to Network Measurements] [Back to NetCalliper Project]

Web Accessibility