The program below is intended to be invoked manually; after printing its task id (obtained with pvm_mytid()), it initiates a copy of another program called hello_other using the pvm_spawn() function. A successful spawn causes the program to execute a blocking receive using pvm_recv. After receiving the message, the program prints the message sent by its counterpart, as well its task id; the buffer is extracted from the message using pvm_upkstr. The final pvm_exit call dissociates the program from the PVM system.
/*
* master.c
*/
#include "pvm3.h"
main()
{
int cc, tid, msgtag;
char buf[100];
printf("i'm t%x\n", pvm_mytid());
cc = pvm_spawn("slave", (char**)0, 0, "", 1, &tid);
if (cc == 1) {
msgtag = 1;
pvm_recv(tid, msgtag);
pvm_upkstr(buf);
printf("from t%x: %s\n", tid, buf);
} else
printf("can't start hello_other\n");
pvm_exit();
}
The program below is a listing of the ``slave'' or spawned program; its first PVM action is to obtain the task id of the ``master'' using the pvm_parent call. This program then obtains its hostname and transmits it to the master using the three-call sequence - pvm_initsend to initialize the send buffer; pvm_pkstr to place a string, in a strongly typed and architecture-independent manner, into the send buffer; and pvm_send to transmit it to the destination process specified by ptid, ``tagging'' the message with the number 1.
/*
* slave.c
*/
#include "pvm3.h"
main()
{
int ptid, msgtag;
char buf[100];
ptid = pvm_parent();
strcpy(buf, "hello, world from ");
gethostname(buf + strlen(buf), 64);
msgtag = 1;
pvm_initsend(PvmDataDefault);
pvm_pkstr(buf);
pvm_send(ptid, msgtag);
pvm_exit();
}
Following the standard PVM convention,
all PVM applications must be located under the
$HOME/pvm3/bin/`/opt/score/pvm/lib/pvmgetarch` directory.
Thus, if this directory has not been created, you must create it as follows:
% mkdir -p $HOME/pvm3/bin/`/opt/score/pvm/lib/pvmgetarch`
To generate a binary and move it to the above directory, the standard
PVM distribution has the aimk tool
which is a wrapper program for make.
The tool set the value of make variables $(PVM_ROOT) and $(PVM_ARCH).
PVM-SCore also supports this tool.
To use aimk, Makefile.aimk is needed
instead of Makefile
Here is a sample Makefile.aimk file:
BDIR = $(HOME)/pvm3/bin
XDIR = $(BDIR)/$(PVM_ARCH)
CFLAGS = -I$(PVM_ROOT)/include
LIBS = -lpvm3 -L$(PVM_ROOT)/lib/$(PVM_ARCH)
all: master slave
master: ../master.c $(XDIR)
$(CC) -o $@ $(CFLAGS) ../$@.c $(LIBS)
mv $@ $(XDIR)
slave: ../slave.c $(XDIR)
$(CC) -o $@ $(CFLAGS) ../$@.c $(LIBS)
mv $@ $(XDIR)
After creating the Makefile.aimk, then simply invokes
the aimk tool as follows:
% aimk all
You may invoke the Compute Host Lock Client,
msgb(1), as follows
to find some free hosts in the cluster:
$ msgb -group pcc &
pcc is a group defined in the
SCore cluster database,
scorehosts.db(5).
Invoke the scout program with the same group
name on which to run your program:
$ scout -g pcc SCOUT: Spawn done. SCOUT: session started $A new shell process is now created as a child of the
scout program.
If msgb is running, some or all hosts in the
msgb window will turn red when the remote processes are
invoked with scout.
Note that the host which the scout program
is invoked doesn't have to be one of the hosts of your cluster.
The following example uses two hosts each of which has a single processor and the maximum number of processes spawned is one.
$ scrun -nodes=2x1 pvmd -e 1 master SCORE: connected (jid=100) <0:0> SCORE: 2 nodes (2x1) ready. ...
/opt/score/pvm/examples/ directory.
/opt/score/pvm/man/ directory.
/opt/score/pvm/doc/release-notes.
/opt/score/pvm/doc/bugreport.
/opt/score/pvm/doc/example.pvmrc.
|
PC Cluster Consotium |