Typical daily throughput

In this subsection the typical data of Wednesday 26/01/2000 and Tuesday 25/01/2000 are presented.

In figure 4 the throughput measurements between the ZAM host and the SARA and UU-36 hosts are given. These hosts are selected here, because all have an interface with a bandwidth of 100 Mbit/s. It is the intension of this plot to check for the speed of the 100 Mbit/s interface at Jülich.

  

Figure: This plot shows the throughput values between the ZAM host and the SARA and the UU-36 hosts. All these hosts have 100 Mbit/s interfaces. Typical data from Wednesday 26/01/2000 and Tuesday 25/01/2000 are presented here. Show plot in real size.

Figure 5 displays the throughput measurements between the IPP host and the FOM and UU-36 hosts. Note that the IPP and FOM hosts have 10 Mbit/s interfaces, in contradiction to the 100 Mbit/s interface of UU-36. This plot is intended to compare the 10 Mbit/s interface of IPP with the 100 Mbit/s interface of ZAM.

  

Figure: This plot shows the throughput values between the IPP host and the FOM and UU-36 hosts. The first two hosts have 10 Mbit/s interfaces, the latter has a 100 Mbit/s interface. Typical data from Wednesday 26/01/2000 and Tuesday 25/01/2000 are presented here. Show plot in real size.

To make a comparison between the two DFN routes to Jülich and Stuttgart, figure 6 shows the throughput data between the FOM host and the IPP (Jülich) and RUS (Stuttgart) hosts. All hosts have 10 Mbit/s interfaces.

  

Figure: This plot displays the throughput values between the FOM host and the IPP and RUS hosts. All hosts have 10 Mbit/s interfaces. Typical data from Wednesday 26/01/2000 and Tuesday 25/01/2000 are presented here. Show plot in real size.

From these figures the following conclusions can be drawn:

1.

In figure 4 we see rather deep daily performance decreases (12:00 - 18:00) especially in the connection ZAM - UU-36. The decrease in the direction ZAM >> UU-36 is deeper and a bit earlier than in the other direction.

2.

The performance decreases (figure 4) are less clear found at in the connection ZAM - UU-36. However, the overall performance of this connection is much lower. The reason therefore is unclear.

3.

In figure 4 we also see performance lost during the evening (19:00 - 23:00). Probably due to Internet browsing from people at home.

4.

In figure 5 the same performance dips appear as in figure 4, however, due to the smaller bandwidth of the interfaces they are less profound.

5.

In figure 5 it also that a higher throughput can be reached at the connection IPP - UU-36, than at IPP - FOM, although the performance dips are deeper and even below the IPP - FOM connection.

6.

Comparing in figure 6 the connection FOM << IPP with FOM << RUS, it appears that the performance in the connections from Jülich, also found in the figures 4 and 5 is less profound in the connection to Stuttgart. However, there is a considerable performance dip in the connection FOM >> RUS.

7.

From all figures it follows that the throughput from the FOM to Germany is lower than in the other direction, while for host UU-36 the reverse is valid.