new simpler approach to parallalization
This commit is contained in:
Florian Stecker
parent
429f0890d6
commit
ac80bc9f3f
@ -1,135 +0,0 @@
|
||||
|
||||
|
||||
euler.ma.utexas.edu
|
||||
fac1.ma.utexas.edu
|
||||
fac4.ma.utexas.edu
|
||||
fac8.ma.utexas.edu
|
||||
fac9.ma.utexas.edu
|
||||
frog.ma.utexas.edu
|
||||
gummo.ma.utexas.edu
|
||||
iguana.ma.utexas.edu
|
||||
lab10.ma.utexas.edu
|
||||
lab11.ma.utexas.edu
|
||||
lab12.ma.utexas.edu
|
||||
lab13.ma.utexas.edu
|
||||
lab14.ma.utexas.edu
|
||||
lab15.ma.utexas.edu
|
||||
lab16.ma.utexas.edu
|
||||
lab17.ma.utexas.edu
|
||||
lab18.ma.utexas.edu
|
||||
lab19.ma.utexas.edu
|
||||
lab1.ma.utexas.edu
|
||||
lab20.ma.utexas.edu
|
||||
lab21.ma.utexas.edu
|
||||
lab22.ma.utexas.edu
|
||||
lab23.ma.utexas.edu
|
||||
lab24.ma.utexas.edu
|
||||
lab25.ma.utexas.edu
|
||||
lab26.ma.utexas.edu
|
||||
lab27.ma.utexas.edu
|
||||
lab28.ma.utexas.edu
|
||||
lab29.ma.utexas.edu
|
||||
lab2.ma.utexas.edu
|
||||
lab30.ma.utexas.edu
|
||||
lab31.ma.utexas.edu
|
||||
lab32.ma.utexas.edu
|
||||
lab33.ma.utexas.edu
|
||||
lab34.ma.utexas.edu
|
||||
lab35.ma.utexas.edu
|
||||
lab36.ma.utexas.edu
|
||||
lab37.ma.utexas.edu
|
||||
lab38.ma.utexas.edu
|
||||
lab39.ma.utexas.edu
|
||||
lab3.ma.utexas.edu
|
||||
lab40.ma.utexas.edu
|
||||
lab41.ma.utexas.edu
|
||||
lab42.ma.utexas.edu
|
||||
lab43.ma.utexas.edu
|
||||
lab44.ma.utexas.edu
|
||||
lab45.ma.utexas.edu
|
||||
lab46.ma.utexas.edu
|
||||
lab47.ma.utexas.edu
|
||||
lab48.ma.utexas.edu
|
||||
lab49.ma.utexas.edu
|
||||
lab4.ma.utexas.edu
|
||||
lab50.ma.utexas.edu
|
||||
lab51.ma.utexas.edu
|
||||
lab52.ma.utexas.edu
|
||||
lab53.ma.utexas.edu
|
||||
lab54.ma.utexas.edu
|
||||
lab55.ma.utexas.edu
|
||||
lab56.ma.utexas.edu
|
||||
lab57.ma.utexas.edu
|
||||
lab58.ma.utexas.edu
|
||||
lab59.ma.utexas.edu
|
||||
lab5.ma.utexas.edu
|
||||
lab60.ma.utexas.edu
|
||||
lab61.ma.utexas.edu
|
||||
lab62.ma.utexas.edu
|
||||
lab63.ma.utexas.edu
|
||||
lab64.ma.utexas.edu
|
||||
lab65.ma.utexas.edu
|
||||
lab66.ma.utexas.edu
|
||||
lab67.ma.utexas.edu
|
||||
lab68.ma.utexas.edu
|
||||
lab69.ma.utexas.edu
|
||||
lab6.ma.utexas.edu
|
||||
lab70.ma.utexas.edu
|
||||
lab7.ma.utexas.edu
|
||||
lab8.ma.utexas.edu
|
||||
lab9.ma.utexas.edu
|
||||
linux100.ma.utexas.edu
|
||||
linux104.ma.utexas.edu
|
||||
linux110.ma.utexas.edu
|
||||
linux115.ma.utexas.edu
|
||||
linux119.ma.utexas.edu
|
||||
linux122.ma.utexas.edu
|
||||
linux149.ma.utexas.edu
|
||||
linux14.ma.utexas.edu
|
||||
linux15.ma.utexas.edu
|
||||
linux164.ma.utexas.edu
|
||||
linux169.ma.utexas.edu
|
||||
linux16.ma.utexas.edu
|
||||
linux17.ma.utexas.edu
|
||||
linux180.ma.utexas.edu
|
||||
linux181.ma.utexas.edu
|
||||
linux184.ma.utexas.edu
|
||||
linux18.ma.utexas.edu
|
||||
linux20.ma.utexas.edu
|
||||
linux21.ma.utexas.edu
|
||||
linux24.ma.utexas.edu
|
||||
linux27.ma.utexas.edu
|
||||
linux28.ma.utexas.edu
|
||||
linux29.ma.utexas.edu
|
||||
linux2.ma.utexas.edu
|
||||
linux30.ma.utexas.edu
|
||||
linux31.ma.utexas.edu
|
||||
linux32.ma.utexas.edu
|
||||
linux38.ma.utexas.edu
|
||||
linux40.ma.utexas.edu
|
||||
linux41.ma.utexas.edu
|
||||
linux46.ma.utexas.edu
|
||||
linux4.ma.utexas.edu
|
||||
linux50.ma.utexas.edu
|
||||
linux52.ma.utexas.edu
|
||||
linux54.ma.utexas.edu
|
||||
linux57.ma.utexas.edu
|
||||
linux62.ma.utexas.edu
|
||||
linux64.ma.utexas.edu
|
||||
linux66.ma.utexas.edu
|
||||
linux68.ma.utexas.edu
|
||||
linux69.ma.utexas.edu
|
||||
linux70.ma.utexas.edu
|
||||
linux71.ma.utexas.edu
|
||||
linux72.ma.utexas.edu
|
||||
linux74.ma.utexas.edu
|
||||
linux76.ma.utexas.edu
|
||||
linux79.ma.utexas.edu
|
||||
linux80.ma.utexas.edu
|
||||
linux82.ma.utexas.edu
|
||||
linux83.ma.utexas.edu
|
||||
linux86.ma.utexas.edu
|
||||
linux91.ma.utexas.edu
|
||||
linux92.ma.utexas.edu
|
||||
linux96.ma.utexas.edu
|
||||
linux9.ma.utexas.edu
|
||||
19
parallelization/generate_commands.py
Executable file
19
parallelization/generate_commands.py
Executable file
@ -0,0 +1,19 @@
|
||||
#!/usr/bin/python
|
||||
|
||||
wordlength = 16
|
||||
n = 101038
|
||||
res = 50
|
||||
radius = 1.0
|
||||
q = [1,1,1]
|
||||
|
||||
denom = round(res/radius)
|
||||
|
||||
cmd = "IDLIST=./output/idlist_{len} ./complex_anosov summary {n} {q1} {q2} {q3} {rnum}/{rden} {inum}/{iden}"
|
||||
|
||||
for i in range(-res,res+1):
|
||||
for j in range(0,res+1):
|
||||
if i == 0 and j == 0:
|
||||
continue
|
||||
print(cmd.format(
|
||||
len=wordlength, n=n, q1=q[0], q2=q[1], q3=q[2],
|
||||
rnum=i, inum=j, rden=denom, iden=denom))
|
||||
@ -1,5 +0,0 @@
|
||||
linux50 slots=4
|
||||
linux52 slots=4
|
||||
linux57 slots=4
|
||||
linux110 slots=4
|
||||
linux115 slots=4
|
||||
@ -1,48 +0,0 @@
|
||||
linux100 slots=4
|
||||
linux104 slots=4
|
||||
linux110 slots=4
|
||||
linux122 slots=4
|
||||
linux149 slots=4
|
||||
linux14 slots=4
|
||||
linux15 slots=2
|
||||
linux16 slots=4
|
||||
linux17 slots=2
|
||||
linux180 slots=4
|
||||
linux181 slots=2
|
||||
linux184 slots=4
|
||||
linux18 slots=4
|
||||
linux20 slots=4
|
||||
linux21 slots=4
|
||||
linux24 slots=4
|
||||
linux27 slots=4
|
||||
linux29 slots=4
|
||||
linux2 slots=4
|
||||
linux30 slots=4
|
||||
linux31 slots=4
|
||||
linux32 slots=4
|
||||
linux38 slots=2
|
||||
linux40 slots=4
|
||||
linux41 slots=4
|
||||
linux46 slots=4
|
||||
linux4 slots=4
|
||||
linux50 slots=4
|
||||
linux52 slots=4
|
||||
linux54 slots=4
|
||||
linux57 slots=4
|
||||
linux62 slots=2
|
||||
linux64 slots=4
|
||||
linux68 slots=4
|
||||
linux69 slots=4
|
||||
linux70 slots=4
|
||||
linux71 slots=4
|
||||
linux72 slots=4
|
||||
linux74 slots=4
|
||||
linux76 slots=4
|
||||
linux79 slots=4
|
||||
linux80 slots=4
|
||||
linux83 slots=4
|
||||
linux86 slots=4
|
||||
linux91 slots=4
|
||||
linux92 slots=4
|
||||
linux96 slots=4
|
||||
linux9 slots=2
|
||||
@ -1,51 +0,0 @@
|
||||
linux100
|
||||
linux104
|
||||
linux110
|
||||
linux115
|
||||
linux122
|
||||
linux149
|
||||
linux14
|
||||
linux15
|
||||
linux164
|
||||
linux169
|
||||
linux16
|
||||
linux17
|
||||
linux180
|
||||
linux181
|
||||
linux184
|
||||
linux18
|
||||
linux20
|
||||
linux21
|
||||
linux24
|
||||
linux27
|
||||
linux29
|
||||
linux2
|
||||
linux30
|
||||
linux31
|
||||
linux32
|
||||
linux38
|
||||
linux40
|
||||
linux41
|
||||
linux46
|
||||
linux4
|
||||
linux50
|
||||
linux52
|
||||
linux54
|
||||
linux57
|
||||
linux62
|
||||
linux64
|
||||
linux68
|
||||
linux69
|
||||
linux70
|
||||
linux71
|
||||
linux72
|
||||
linux74
|
||||
linux76
|
||||
linux79
|
||||
linux80
|
||||
linux83
|
||||
linux86
|
||||
linux91
|
||||
linux92
|
||||
linux96
|
||||
linux9
|
||||
@ -1,409 +0,0 @@
|
||||
#include "parallel.h"
|
||||
|
||||
#include <mpi.h>
|
||||
#include <sys/stat.h>
|
||||
#include <sys/mman.h>
|
||||
#include <fcntl.h>
|
||||
#include <errno.h>
|
||||
#include <string.h>
|
||||
#include <unistd.h>
|
||||
#include <malloc.h>
|
||||
#include <stdlib.h>
|
||||
|
||||
//#define DEBUG INFO
|
||||
#define DEBUG(msg, ...)
|
||||
#define INFO(msg, ...) fprintf(stderr, "[%003d%10.3f] " msg, mpi_rank(0), runtime(), ##__VA_ARGS__)
|
||||
//#define DEBUG(msg, ...) fprintf(stderr, "[ %10.3f] " msg, runtime(), ##__VA_ARGS__)
|
||||
//#define DEBUG_MPI(msg, node, ...) fprintf(stderr, "[%003d%10.3f] " msg, node, runtime(), ##__VA_ARGS__)
|
||||
#define DONE(x) *((int*)(x))
|
||||
|
||||
enum message_tag {
|
||||
PARALLEL_ORDER = 0,
|
||||
PARALLEL_RESULT,
|
||||
PARALLEL_SHUTDOWN,
|
||||
PARALLEL_GLOBAL_DATA
|
||||
};
|
||||
|
||||
struct timespec starttime;
|
||||
|
||||
int mpi_rank(int activate_mpi)
|
||||
{
|
||||
static int active = 0;
|
||||
if(activate_mpi)
|
||||
active = 1;
|
||||
|
||||
if(!active)
|
||||
return 0;
|
||||
else {
|
||||
int rank;
|
||||
MPI_Comm_rank(MPI_COMM_WORLD, &rank);
|
||||
return rank;
|
||||
}
|
||||
}
|
||||
|
||||
void start_timer()
|
||||
{
|
||||
clock_gettime(CLOCK_MONOTONIC, &starttime);
|
||||
}
|
||||
|
||||
double runtime()
|
||||
{
|
||||
struct timespec curtime;
|
||||
double diff;
|
||||
clock_gettime(CLOCK_MONOTONIC, &curtime);
|
||||
return (curtime.tv_sec - starttime.tv_sec) + (curtime.tv_nsec - starttime.tv_nsec) / 1e9;
|
||||
}
|
||||
|
||||
parallel_context *parallel_init()
|
||||
{
|
||||
parallel_context *ctx = malloc(sizeof(parallel_context));
|
||||
|
||||
if(!getenv("OMPI_COMM_WORLD_SIZE")) {
|
||||
ctx->mpi_mode = 0;
|
||||
DEBUG("Running standalone.\n");
|
||||
return ctx;
|
||||
}
|
||||
|
||||
ctx->mpi_mode = 1;
|
||||
int result = MPI_Init(NULL, NULL);
|
||||
MPI_Comm_size(MPI_COMM_WORLD, &ctx->size);
|
||||
MPI_Comm_rank(MPI_COMM_WORLD, &ctx->rank);
|
||||
MPI_Get_processor_name(ctx->processor_name, &ctx->processor_name_len);
|
||||
|
||||
mpi_rank(1); // display the rank in debug output from now on
|
||||
|
||||
if(ctx->rank == 0)
|
||||
DEBUG("Running in mpi mode, %d nodes.\n", ctx->size);
|
||||
|
||||
return ctx;
|
||||
}
|
||||
|
||||
void parallel_destroy(parallel_context* ctx)
|
||||
{
|
||||
if(ctx->mpi_mode) {
|
||||
MPI_Type_free(&ctx->order_datatype);
|
||||
MPI_Type_free(&ctx->result_datatype);
|
||||
MPI_Finalize();
|
||||
}
|
||||
|
||||
free(ctx);
|
||||
}
|
||||
|
||||
void parallel_set_datasize_and_callbacks(parallel_context *ctx, parallel_callback_init init, parallel_callback_job job, parallel_callback_destroy destroy, int global_data_size, int node_data_size, int input_size, int output_size)
|
||||
{
|
||||
ctx->init = init;
|
||||
ctx->destroy = destroy;
|
||||
ctx->job = job;
|
||||
ctx->global_data_size = global_data_size;
|
||||
ctx->node_data_size = node_data_size;
|
||||
ctx->input_size = input_size;
|
||||
ctx->output_size = output_size;
|
||||
|
||||
if(ctx->mpi_mode) {
|
||||
// create a datatype for job orders, consisting of an integer (the job id) and a user-defined section
|
||||
int order_blocklengths[2] = {1, input_size};
|
||||
MPI_Aint order_displacements[2] = {0, sizeof(int)};
|
||||
MPI_Datatype order_types[2] = {MPI_INT, MPI_BYTE};
|
||||
MPI_Type_create_struct(2, order_blocklengths, order_displacements, order_types, &ctx->order_datatype);
|
||||
MPI_Type_commit(&ctx->order_datatype);
|
||||
|
||||
int result_blocklengths[2] = {1, output_size};
|
||||
MPI_Aint result_displacements[2] = {0, sizeof(int)};
|
||||
MPI_Datatype result_types[2] = {MPI_INT, MPI_BYTE};
|
||||
MPI_Type_create_struct(2, result_blocklengths, result_displacements, result_types, &ctx->result_datatype);
|
||||
MPI_Type_commit(&ctx->result_datatype);
|
||||
}
|
||||
}
|
||||
|
||||
int parallel_job(parallel_context *ctx, const void *global_data, void *node_data, int block)
|
||||
{
|
||||
MPI_Status status;
|
||||
double jobtime;
|
||||
void *input_and_job_nr = malloc(ctx->input_size + sizeof(int));
|
||||
void *output_and_job_nr = malloc(ctx->output_size + sizeof(int));
|
||||
void *input = input_and_job_nr + sizeof(int);
|
||||
int *job_nr = (int *)input_and_job_nr;
|
||||
void *output = output_and_job_nr + sizeof(int);
|
||||
int *output_job_nr = (int *)output_and_job_nr;
|
||||
int result = 0;
|
||||
int message_present;
|
||||
|
||||
if(block) {
|
||||
jobtime = -MPI_Wtime();
|
||||
MPI_Probe(0, MPI_ANY_TAG, MPI_COMM_WORLD,
|
||||
&status);
|
||||
jobtime += MPI_Wtime();
|
||||
|
||||
INFO("TIMING: Probe() took %f seconds\n", jobtime);
|
||||
message_present = 1;
|
||||
} else {
|
||||
MPI_Iprobe(0, MPI_ANY_TAG, MPI_COMM_WORLD,
|
||||
&message_present, &status);
|
||||
}
|
||||
|
||||
// DEBUG("Message received: source = %d, tag = %d\n", status.MPI_SOURCE, status.MPI_TAG);
|
||||
|
||||
if(message_present) {
|
||||
if(status.MPI_TAG == PARALLEL_SHUTDOWN) {
|
||||
DEBUG("Shutting down\n");
|
||||
result = 1;
|
||||
} else if(status.MPI_TAG == PARALLEL_ORDER) {
|
||||
MPI_Recv(input_and_job_nr,
|
||||
1, ctx->order_datatype,
|
||||
0, PARALLEL_ORDER, MPI_COMM_WORLD,
|
||||
&status);
|
||||
|
||||
DEBUG("Working on job %d\n", *job_nr);
|
||||
|
||||
jobtime = -MPI_Wtime();
|
||||
|
||||
// do the actual work
|
||||
ctx->job(global_data, node_data, input, output);
|
||||
|
||||
jobtime += MPI_Wtime();
|
||||
|
||||
INFO("TIMING: job %d took %f seconds\n", *job_nr, jobtime);
|
||||
|
||||
*output_job_nr = *job_nr;
|
||||
jobtime = -MPI_Wtime();
|
||||
MPI_Send(output_and_job_nr,
|
||||
1, ctx->result_datatype,
|
||||
0, PARALLEL_RESULT, MPI_COMM_WORLD);
|
||||
jobtime += MPI_Wtime();
|
||||
|
||||
INFO("TIMING: Send() took %f seconds\n", jobtime);
|
||||
}
|
||||
} else {
|
||||
result = 2;
|
||||
}
|
||||
|
||||
free(input_and_job_nr);
|
||||
free(output_and_job_nr);
|
||||
|
||||
return result;
|
||||
}
|
||||
|
||||
int parallel_work(parallel_context *ctx)
|
||||
{
|
||||
// do nothing in non-mpi mode
|
||||
if(ctx->mpi_mode == 0)
|
||||
return 0;
|
||||
|
||||
void *global_data = malloc(ctx->global_data_size);
|
||||
void *node_data = malloc(ctx->node_data_size);
|
||||
|
||||
// wait for global data
|
||||
MPI_Bcast(global_data, ctx->global_data_size, MPI_BYTE, 0, MPI_COMM_WORLD);
|
||||
|
||||
DEBUG("Global data received\n");
|
||||
|
||||
// initialize node_data (and do once-per-node computation)
|
||||
ctx->init(global_data, node_data);
|
||||
|
||||
DEBUG("Initialization completed\n");
|
||||
|
||||
int shutdown = 0;
|
||||
while(shutdown != 1) {
|
||||
shutdown = parallel_job(ctx, global_data, node_data, 1);
|
||||
}
|
||||
|
||||
ctx->destroy(global_data, node_data);
|
||||
|
||||
free(global_data);
|
||||
free(node_data);
|
||||
|
||||
return 0;
|
||||
}
|
||||
|
||||
int parallel_run(parallel_context *ctx, const void *global_data, const void *input_array, void *output_array, unsigned int njobs, const char *_restart_filename)
|
||||
{
|
||||
// in non-mpi-mode, just run init, forall(jobs) job
|
||||
if(ctx->mpi_mode == 0) {
|
||||
void *node_data = malloc(ctx->node_data_size);
|
||||
int result = ctx->init(global_data, node_data);
|
||||
if(result != 0)
|
||||
goto cleanup_standalone;
|
||||
|
||||
for(int i = 0; i < njobs; i++) {
|
||||
result = ctx->job(
|
||||
global_data,
|
||||
node_data,
|
||||
input_array + ctx->input_size*i,
|
||||
output_array + ctx->output_size*i);
|
||||
if(result != 0)
|
||||
goto cleanup_standalone;
|
||||
}
|
||||
|
||||
cleanup_standalone:
|
||||
ctx->destroy(global_data, node_data);
|
||||
return result;
|
||||
} else {
|
||||
// if no restart file was specified, pick a filename
|
||||
char *restart_filename;
|
||||
char buffer[128];
|
||||
int restartf;
|
||||
if(_restart_filename == NULL) {
|
||||
time_t t = time(NULL);
|
||||
struct tm *loctm = localtime(&t);
|
||||
strftime(buffer, sizeof(buffer), "restart/restart_%y%m%d_%H%M%S", loctm);
|
||||
restart_filename = buffer;
|
||||
} else {
|
||||
restart_filename = (char *)_restart_filename;
|
||||
}
|
||||
|
||||
// open restart file if it exists, otherwise create it
|
||||
int continuing = 1;
|
||||
restartf = open(restart_filename, O_RDWR);
|
||||
if(restartf == -1 && errno == ENOENT) {
|
||||
restartf = open(restart_filename, O_RDWR | O_CREAT, 0666);
|
||||
continuing = 0;
|
||||
}
|
||||
if(restartf == -1) {
|
||||
DEBUG("Error opening restart file: %s\n", strerror(errno));
|
||||
exit(1);
|
||||
}
|
||||
|
||||
// map restart file
|
||||
int itemsize = (ctx->output_size + sizeof(int)); // for every job, store output, and completed flag
|
||||
ftruncate(restartf, njobs*itemsize);
|
||||
void *alljobs = mmap(0, njobs*itemsize, PROT_READ | PROT_WRITE, MAP_SHARED, restartf, 0);
|
||||
if(alljobs == MAP_FAILED) {
|
||||
DEBUG("Error mapping restart file: %s\n", strerror(errno));
|
||||
exit(1);
|
||||
}
|
||||
|
||||
// count completed jobs, or initialize jobs
|
||||
int completed = 0;
|
||||
if(continuing) {
|
||||
for(int i = 0; i < njobs; i++)
|
||||
if(DONE(alljobs + i*itemsize))
|
||||
completed++;
|
||||
} else {
|
||||
for(int i = 0; i < njobs; i++) {
|
||||
DONE(alljobs + i*itemsize) = 0;
|
||||
memcpy(alljobs + i*itemsize + sizeof(int), input_array + i*ctx->input_size, ctx->input_size); // copy input data
|
||||
}
|
||||
}
|
||||
|
||||
fsync(restartf);
|
||||
|
||||
if(continuing) {
|
||||
INFO("Continuing from restart file, %d/%d jobs completed, %d nodes\n", completed, njobs, ctx->size);
|
||||
} else {
|
||||
INFO("Starting from scratch, %d jobs, %d nodes\n", njobs, ctx->size);
|
||||
}
|
||||
|
||||
if(completed >= njobs)
|
||||
goto cleanup_mpi;
|
||||
|
||||
/* Send global data */
|
||||
MPI_Bcast((void*)global_data, ctx->global_data_size, MPI_BYTE, 0, MPI_COMM_WORLD);
|
||||
|
||||
DEBUG("Global data sent\n");
|
||||
|
||||
// we want to be able to run jobs ourselves, so initialize node_data
|
||||
void *node_data = malloc(ctx->node_data_size);
|
||||
ctx->init(global_data, node_data);
|
||||
|
||||
void *input_message_buffer = malloc(ctx->input_size + sizeof(int));
|
||||
void *output_message_buffer = malloc(ctx->output_size + sizeof(int));
|
||||
int *active_jobs = malloc(sizeof(int)*ctx->size);
|
||||
memset(active_jobs, 0, ctx->size*sizeof(int));
|
||||
int active_worker_nodes = ctx->size - 1; // we don't count ourselves, since we can't shut ourselves down
|
||||
|
||||
// find next unfinished job
|
||||
int current = 0;
|
||||
while(current < njobs && DONE(alljobs + current*itemsize))
|
||||
current++;
|
||||
|
||||
// assign initial jobs, 2 for each worker thread
|
||||
for(int i = 0; i < 2*ctx->size; i++) {
|
||||
if(current >= njobs) // all jobs are assigned
|
||||
break;
|
||||
|
||||
// send job id and input data
|
||||
// send to all nodes except ourself (node 0)
|
||||
*((int*)input_message_buffer) = current;
|
||||
memcpy(input_message_buffer + sizeof(int), input_array + current*ctx->input_size, ctx->input_size);
|
||||
MPI_Send(input_message_buffer, 1, ctx->order_datatype,
|
||||
i%ctx->size, PARALLEL_ORDER, MPI_COMM_WORLD);
|
||||
|
||||
DEBUG("Job %d sent to node %d\n", current, i%ctx->size);
|
||||
active_jobs[i%ctx->size]++;
|
||||
current++;
|
||||
}
|
||||
|
||||
MPI_Status status;
|
||||
int message_present;
|
||||
while(active_jobs[0] != 0 || active_worker_nodes != 0) {
|
||||
MPI_Iprobe(MPI_ANY_SOURCE, PARALLEL_RESULT, MPI_COMM_WORLD, &message_present, &status);
|
||||
DEBUG("Message present, tag = %d, source = %d\n", status.MPI_TAG, status.MPI_SOURCE);
|
||||
if(!message_present) {
|
||||
// if there are no more messages to process, we can run a job ourselves before returning to managing
|
||||
DEBUG("Start running job myself\n");
|
||||
int result = parallel_job(ctx, global_data, node_data, 0);
|
||||
DEBUG("Finished running job myself, result = %d\n");
|
||||
} else if(status.MPI_TAG == PARALLEL_RESULT) {
|
||||
MPI_Recv(output_message_buffer, 1, ctx->result_datatype,
|
||||
MPI_ANY_SOURCE, PARALLEL_RESULT, MPI_COMM_WORLD, &status);
|
||||
|
||||
DEBUG("Got message tag %d from node %d\n", status.MPI_TAG, status.MPI_SOURCE);
|
||||
|
||||
int node = status.MPI_SOURCE;
|
||||
int id = *((int*)output_message_buffer);
|
||||
memcpy(alljobs + id*itemsize + sizeof(int), output_message_buffer + sizeof(int), ctx->output_size);
|
||||
DONE(alljobs + id*itemsize) = 1;
|
||||
active_jobs[node]--;
|
||||
|
||||
// todo: deal with unresponsive nodes
|
||||
// strategy: when no jobs left, go through unfinished list again, incrementing oversubscribe counter
|
||||
// if oversubscribe counter is at limit, shut node down instead
|
||||
//
|
||||
|
||||
if(current >= njobs) { // all jobs are assigned, try to shut down node
|
||||
// don't try to shut down ourselves, and only if it has no other jobs to do
|
||||
if(node != 0 && active_jobs[node] == 0) {
|
||||
MPI_Send(NULL, 0, MPI_BYTE, node, PARALLEL_SHUTDOWN, MPI_COMM_WORLD);
|
||||
active_worker_nodes--;
|
||||
INFO("job %d completed by node %d, shut down, %d workers remaining\n", id, node, active_worker_nodes);
|
||||
}
|
||||
} else {
|
||||
*((int*)input_message_buffer) = current;
|
||||
memcpy(input_message_buffer + sizeof(int), input_array + current*ctx->input_size, ctx->input_size);
|
||||
MPI_Send(input_message_buffer, 1, ctx->order_datatype,
|
||||
node, PARALLEL_ORDER, MPI_COMM_WORLD);
|
||||
active_jobs[node]++;
|
||||
current++;
|
||||
|
||||
if(active_jobs[node] < 3) {
|
||||
*((int*)input_message_buffer) = current;
|
||||
memcpy(input_message_buffer + sizeof(int), input_array + current*ctx->input_size, ctx->input_size);
|
||||
MPI_Send(input_message_buffer, 1, ctx->order_datatype,
|
||||
node, PARALLEL_ORDER, MPI_COMM_WORLD);
|
||||
active_jobs[node]++;
|
||||
current++;
|
||||
|
||||
INFO("job %d completed by node %d, continues with %d and %d\n", id, node, current-1, current-2);
|
||||
} else {
|
||||
INFO("job %d completed by node %d, continues with %d\n", id, node, current-1);
|
||||
}
|
||||
}
|
||||
}
|
||||
}
|
||||
|
||||
for(int i = 0; i < njobs; i++) {
|
||||
memcpy(output_array + i*ctx->output_size, alljobs + i*itemsize + sizeof(int), ctx->output_size);
|
||||
}
|
||||
|
||||
free(input_message_buffer);
|
||||
free(output_message_buffer);
|
||||
free(node_data);
|
||||
free(active_jobs);
|
||||
|
||||
cleanup_mpi:
|
||||
munmap(alljobs, njobs*itemsize);
|
||||
close(restartf);
|
||||
}
|
||||
|
||||
return 0;
|
||||
}
|
||||
@ -1,118 +0,0 @@
|
||||
#ifndef PARALLEL_H
|
||||
#define PARALLEL_H
|
||||
|
||||
/*
|
||||
this is a library to parallelize workloads which can be split up naturally
|
||||
into a sequence of independent jobs, using MPI. A program will usually
|
||||
|
||||
- do precomputation
|
||||
- fill array with input data
|
||||
- do the parallel work
|
||||
- print the output data
|
||||
|
||||
we want to enable restarts, so that only unfinshed jobs need to be repeated.
|
||||
Further, we want to be resilient to slow/unreliable network and to losing
|
||||
nodes. There is a main node and a number of workers. The main node does the
|
||||
precomputation and then retires do do administrative work, and the workers
|
||||
do the actual jobs. We also want to switch to serial mode if the program is
|
||||
called without MPI.
|
||||
|
||||
The following data has to be transimitted between nodes:
|
||||
- results of the precomputation (read-only, shared between nodes)
|
||||
- job-specific input data, generated by main node before parallel part
|
||||
- output data for each job
|
||||
|
||||
the parallel work shall be given as a callback function which takes
|
||||
input data and precomputation data as parameter
|
||||
|
||||
the above program will look like this for us:
|
||||
|
||||
- parallel_init
|
||||
- if we are a worker, do parallel_work(init_callback, job_callback), exit
|
||||
- do precomputation
|
||||
- fill array with input data
|
||||
- output_array = parallel_run(input_array)
|
||||
- print the output data
|
||||
|
||||
parallel_init:
|
||||
- check if we're running as an mpi program
|
||||
- init mpi, check what kind of node we are
|
||||
|
||||
parallel_work(init_callback1, init_callback2, job_callback):
|
||||
- receive global_precomp (???)
|
||||
- worker_precomp = init_callback2(global_precomp, worker_precomp)
|
||||
- infinite loop:
|
||||
- wait for job on network, receive input
|
||||
- output = job_callback(global_precomp, worker_precomp, input)
|
||||
- send output on network
|
||||
- exit loop on shutdown signal
|
||||
|
||||
parallel_run(global_precomp, input_array, restart file, callbacks):
|
||||
- check if we're running as an MPI program
|
||||
- send global_precomp to all nodes (if MPI)
|
||||
- if(restart file given and exists) read restart file
|
||||
- else create new restart file
|
||||
- until(all jobs finished):
|
||||
- if MPI:
|
||||
- send next job & input to appropriate node
|
||||
- if all jobs are in work, reassign unfinished ones (up to limit)
|
||||
- collect outputs
|
||||
- if no MPI:
|
||||
- worker_precomp = init_callback1
|
||||
- worker_precomp = init_callback2(global_precomp, worker_precomp)
|
||||
- for(j in jobs)
|
||||
- output(j) = job_callback(global_precomp, worker_precomp, input(j))
|
||||
- delete restart file
|
||||
- return array of outputs
|
||||
|
||||
parallel_destroy():
|
||||
- free everything
|
||||
|
||||
have a context? probably yes: parallel_context
|
||||
|
||||
plan:
|
||||
- make interface
|
||||
- implement no-MPI part
|
||||
- restructure singular_values.c to use interface
|
||||
- implement MPI part
|
||||
*/
|
||||
|
||||
#include <mpi.h>
|
||||
#include <time.h>
|
||||
|
||||
typedef void (*parallel_callback_destroy)(const void*, void*);
|
||||
typedef int (*parallel_callback_init)(const void*,void*);
|
||||
typedef int (*parallel_callback_job)(const void*,void*,const void*,void*);
|
||||
|
||||
typedef struct {
|
||||
int mpi_mode;
|
||||
struct timespec starttime;
|
||||
char processor_name[MPI_MAX_PROCESSOR_NAME];
|
||||
int processor_name_len;
|
||||
int rank;
|
||||
int size;
|
||||
MPI_Datatype order_datatype;
|
||||
MPI_Datatype result_datatype;
|
||||
parallel_callback_init init;
|
||||
parallel_callback_job job;
|
||||
parallel_callback_destroy destroy;
|
||||
void *global_data;
|
||||
void *node_data;
|
||||
int global_data_size;
|
||||
int node_data_size;
|
||||
int input_size;
|
||||
int output_size;
|
||||
} parallel_context;
|
||||
|
||||
parallel_context *parallel_init();
|
||||
void parallel_set_datasize_and_callbacks(parallel_context *ctx, parallel_callback_init init, parallel_callback_job job, parallel_callback_destroy destroy, int global_data_size, int node_data_size, int input_size, int output_size);
|
||||
int parallel_work(parallel_context *ctx);
|
||||
int parallel_run(parallel_context *ctx, const void *global_data, const void *input_array, void *output_array, unsigned int njobs, const char *restart_filename);
|
||||
void parallel_destroy(parallel_context* ctx);
|
||||
|
||||
int mpi_rank();
|
||||
void start_timer();
|
||||
double runtime();
|
||||
|
||||
|
||||
#endif
|
||||
@ -1,9 +0,0 @@
|
||||
#!/bin/bash
|
||||
|
||||
# nmax=895882 # up to reflection group word length 22 ( 555 group)
|
||||
nmax=700000 # up to reflection group word length 22 ( 444 group)
|
||||
# nmax=11575 # up to reflection group word length 14
|
||||
|
||||
#time mpirun --mca opal_warn_on_missing_libcuda 0 -x LD_LIBRARY_PATH=/home/stecker/svmpi/libs ./singular_values $nmax ejp_trg_restart test.out
|
||||
|
||||
time mpirun --mca opal_warn_on_missing_libcuda 0 --mca mpi_yield_when_idle 1 -np 4 ./singular_values 700000 4 4 4 1 1 100 1 100 100 $1
|
||||
@ -1,14 +0,0 @@
|
||||
#!/bin/bash
|
||||
|
||||
cd /home/stecker/svmpi/
|
||||
|
||||
nmax=895882 # up to reflection group word length 22
|
||||
# nmax=11575 # up to reflection group word length 14
|
||||
|
||||
outfile=result_$(date +%Y%m%d_%H%M%S).out
|
||||
|
||||
|
||||
unset DISPLAY
|
||||
|
||||
make singular_values &&
|
||||
time mpirun -n 100 -x LD_LIBRARY_PATH=/home/stecker/svmpi/libs --hostfile hostfile_big ./singular_values $nmax utexas_cluster_restart $outfile
|
||||
57
parallelization/runjobs.py
Executable file
57
parallelization/runjobs.py
Executable file
@ -0,0 +1,57 @@
|
||||
#!/usr/bin/python
|
||||
|
||||
from mpi4py import MPI
|
||||
import os
|
||||
import re
|
||||
import math
|
||||
import subprocess
|
||||
import time
|
||||
|
||||
comm = MPI.COMM_WORLD
|
||||
rank = comm.Get_rank()
|
||||
nodes = comm.Get_size()
|
||||
|
||||
# print(os.path.abspath(os.curdir))
|
||||
|
||||
done = set()
|
||||
for f in os.listdir('.'):
|
||||
if re.search('^done_[0-9]+', f):
|
||||
fp = open(f, "r")
|
||||
for x in fp:
|
||||
done.add(int(x))
|
||||
fp.close()
|
||||
|
||||
f = open("commands", "r")
|
||||
idx = 0
|
||||
todo = []
|
||||
for c in f:
|
||||
if not idx in done:
|
||||
todo.append((idx,c))
|
||||
idx = idx+1
|
||||
f.close()
|
||||
|
||||
start = math.floor(len(todo)/nodes*rank)
|
||||
end = math.floor(len(todo)/nodes*(rank+1))
|
||||
if(rank == nodes-1):
|
||||
end = len(todo)
|
||||
|
||||
print("{n:d} commands awaiting execution, {nnode:d} of them in node {rank:d}".format(n=len(todo),nnode=end-start,rank=rank))
|
||||
|
||||
time.sleep(1) # to make sure all nodes read the status first before more gets done
|
||||
|
||||
outfilename = "result_{node:003d}".format(node=rank)
|
||||
donefilename = "done_{node:003d}".format(node=rank)
|
||||
outfile = open(outfilename, "a")
|
||||
donefile = open(donefilename, "a")
|
||||
|
||||
for i in range(start, end):
|
||||
result = subprocess.call(todo[i][1], stdout=outfile, shell=True)
|
||||
if result == 0:
|
||||
donefile.write(str(todo[i][0]) + '\n')
|
||||
else:
|
||||
print("Command failed: {cmd}".format(cmd=todo[i][1]))
|
||||
outfile.flush()
|
||||
donefile.flush()
|
||||
|
||||
outfile.close()
|
||||
donefile.close()
|
||||
@ -1,20 +0,0 @@
|
||||
#!/bin/bash
|
||||
|
||||
#SBATCH -J ejp_trg
|
||||
#SBATCH -o logs/ejp_trg.o%j
|
||||
#SBATCH -e logs/ejp_trg.e%j
|
||||
#SBATCH -p skx-dev
|
||||
#SBATCH -N 1
|
||||
#SBATCH -n 48
|
||||
#SBATCH -t 00:05:00
|
||||
#SBATCH --mail-user=mail@florianstecker.net
|
||||
#SBATCH --mail-type=all
|
||||
|
||||
export LD_LIBRARY_PATH=$WORK/mps/lib:$LD_LIBRARY_PATH
|
||||
|
||||
d=$(date +%Y%m%d_%H%M%S)
|
||||
|
||||
nmax=895882 # up to reflection group word length 22
|
||||
# nmax=11575 # up to reflection group word length 1
|
||||
|
||||
ibrun ./singular_values $nmax $SCRATCH/ejp_trg_restart $WORK/ejp_trg/output/result_$d
|
||||
@ -1,8 +0,0 @@
|
||||
#!/bin/bash
|
||||
|
||||
rsync -vt *.c *.h Makefile stampede.slurm stampede:work/ejp_trg/
|
||||
#rsync -lvt /usr/lib/libmps.so* /usr/include/mps utexas:work/ejp_trg/libs/
|
||||
|
||||
# now run it with a job script
|
||||
|
||||
# get MPSolve from https://numpi.dm.unipi.it/_media/software/mpsolve/mpsolve-3.2.1.tar.bz2
|
||||
@ -1,7 +0,0 @@
|
||||
#!/bin/bash
|
||||
|
||||
rsync -vt *.c *.h Makefile hostfile hostfile_big allhosts localnames run_utexas run_local utexas:svmpi/
|
||||
rsync -lvt /usr/lib/libmps.so* utexas:svmpi/libs/
|
||||
rsync -rvt /usr/include/mps utexas:svmpi/libs/
|
||||
|
||||
# now run it with ssh utexas -t ssh linux50 svmpi/run_utexas
|
||||
Loading…
Reference in New Issue
Block a user