In reference to the HEC meeting to be held in the last week of May 2011, we tried to analyse the network performance for PoP node (pingerjms.pern.edu.pk) and Non PoP node (pinger.usindh.edu.pk) both at University of Sindh Jamshoro.
Metric Measurements and Results
Looking at the pingtable.pl and getting data from Pakistan to pingerjms and Pakistan to usindh node, we see the following results that show the performance of both of the nodes in the first 15 days of May.
Find spreadsheet here.
The graphs clearly show that the PoP node performance is worse than the non-PoP node. In order to get a better insight into finding the reason of this degraded performance, we looked at the traceroutes for both of the nodes from pinger.pern.edu.pk (Islamabad Pern PoP). However, the traceroutes for both of the nodes are throwing error messages.
Traceroute Results from pinger.pern.edu.pk
While trying to access pingerjms.pern.edu.pk (PoP Jamshoro) from pinger.pern.edu.pk (PoP Islamabad) following traceroute results were obtained
Executing exec(traceroute, -m 30 -q 3 -f 3, 121.52.154.2, 140) traceroute to 121.52.154.2 (121.52.154.2), 30 hops max, 140 byte packets 4 rwp44.pie.net.pk (221.120.251.58) 35.530 ms 35.491 ms 35.424 ms 5 khi77.pie.net.pk (221.120.205.70) 36.129 ms 36.417 ms 36.490 ms 6 pinger.hyd.usindh.edu.pk (121.52.154.2) 35.945 ms 35.931 ms 35.916 ms
However, while accessing usindh (non PoP Jamshoro) we found
Executing exec(traceroute, -m 30 -q 3 -f 3, 121.52.155.62, 140) traceroute to 121.52.155.62 (121.52.155.62), 30 hops max, 140 byte packets 4 rwp44.pie.net.pk (221.120.251.58) 35.473 ms 35.442 ms 35.451 ms 5 khi77.pie.net.pk (221.120.205.70) 36.107 ms 36.347 ms 36.388 ms 6 172.31.253.66 (172.31.253.66) 36.077 ms 36.201 ms 36.248 ms 7 usindh.edu.pk (121.52.156.130) 36.550 ms 36.583 ms 36.623 ms 8 (121.52.155.62) 36.917 ms !X 36.899 ms !X 36.885 ms !X
Extra hops can be seen in Usindh traceroute (hop 6 and 7). For both of the nodes, it takes 36.105 ~ 36.4 ms to reach the hop khi77.pie.net.pk, however, as the hops are increased for accessing usindh , an increased RTT of 0.9 ms is observed. This is a small difference in RTT for the two nodes. This difference can be due to temporary congested network and long queues.