Grid-Enabling Applications with JGRIM - Cloud, Grid and High Performance Computing: Emerging Applications - page 51

Figure 4. Grid used for the experiments

Table 1. Hardware specifications of the Grid machines

Machine
name

Memory
(MB)

Cluster name CPU model CPU frequency

A.1 AMD Athlon XP 2200+ 1.75 Ghz. 256

A

A.2 Intel Core2 T5600 1.83 Ghz. (per core) 1.024

B.1 AMD Sempron 1.90 Ghz. 512

B

B.2 AMD Athlon 64 X2 Dual Core 3.600+ 2.00 Ghz. (per core) 1.024

C.1 Intel Pentium 4 2.80 Ghz. 512

C.2 Intel Pentium III (Coppermine) 852 Mhz. 256

C.3 Intel Pentium III (Coppermine) 852 Mhz. 256

C

C.4 Intel Pentium III (Coppermine) 852 Mhz. 384

C.5 Intel Pentium III (Coppermine) 852 Mhz. 384

C.6 Intel Pentium III (Coppermine) 798 Mhz. 256

but minimal background on JGRIM, ProActive
and Satin. All experiments were performed dur-
ing nighttime (from 11 P.M. to 8 A.M.), when the
Internet traffic is low and the network latency has
little variability.
Table 1 details the CPU and memory specifi-
cations of the nodes of the previous Grid setting.

Machines were equipped with Ubuntu Linux
(kernel version 2.6.20) and the Sun JDK 1.5.0.
The reason of using such an heterogeneous hard-
ware was to establish a realistic Grid testbed for
the experiments.
We assessed the impact of gridification on the
application code when employing the three tools