\section{Methodology}

Our approach is to directly measure the network traffic of each of the
programs on a LAN of Ethernet~\cite{METCALFE-ETHERNET} connected DEC
Alpha~\cite{DEC-3000-JOURNAL} workstations.  A machine running in
promiscuous mode is used to record each packet.  This data is then
analyzed using a variety of simple, custom programs.


\subsection{Environment}

Nine DEC 3000/400 Alpha (21064~\cite{DEC-ALPHA-21064-JOURNAL} at 133
MHz with 64 MB RAM) workstations~\cite{DEC-3000-JOURNAL} running OSF/1
2.0 were used as our testbed.  The built-in
Ethernet~\cite{METCALFE-ETHERNET} adaptors were married to a
multi-segment bridged Ethernet LAN, so all machines shared a common
collision domain and an aggregate 1.25 MB/s of bandwidth.  Since these
machines are office workstations and other machines share the LAN, all
measurements were performed in the early morning hours (4-5 am) to
avoid other traffic, and were repeated several times.


\subsection{Compilation}

Each of the six Fx programs can be compiled for an arbitrary number of
processors.  Due to the stress these programs place on machines and
networks, it was decided to compile them for four processors.
The programs were compiled with version 2.2 of Fx compiler and version
3.3 of the DEC Fortran compiler.  The basic level of optimization
(-O) was used with the latter compiler.  The object files were linked
with version 3.3.3 of PVM and with version 2.2 of the Fx/PVM run-time
system.

\subsection{Measurement}

To measure the network traffic, one of the workstations was configured
with the DEC packet filter software, which allows priveledged users to
use the network adaptor in promiscuous mode.  The measurement
workstation was not used to run any Fx program.  Instead, it ran the
TCPDUMP program included with OSF/1 and collected a trace of all the
packets on the LAN generated by each test program.  For the Fx
programs, including AIRSHED, each outer loop as iterated 100 times,
except for SEQ, which was iterated five times.

Each of our traces captured all the packets on the network, providing
a time stamp, size, protocol, source and destination for each packet.
We considered the size of the packet to include the data portion, TCP or
UDP header, IP header, and Ethernet header and trailer.  Where sensible,
we produced a trace for a single connection by extracting all packets
sent from one host to another.