1
1
openmpi/orte/mca/pls/tm/pls_tm_module.c
Ralph Castain 85df3bd92f Bring in the generalized xcast communication system along with the correspondingly revised orted launch. I will send a message out to developers explaining the basic changes. In brief:
1. generalize orte_rml.xcast to become a general broadcast-like messaging system. Messages can now be sent to any tag on the daemons or processes. Note that any message sent via xcast will be delivered to ALL processes in the specified job - you don't get to pick and choose. At a later date, we will introduce an augmented capability that will use the daemons as relays, but will allow you to send to a specified array of process names.

2. extended orte_rml.xcast so it supports more scalable message routing methodologies. At the moment, we support three: (a) direct, which sends the message directly to all recipients; (b) linear, which sends the message to the local daemon on each node, which then relays it to its own local procs; and (b) binomial, which sends the message via a binomial algo across all the daemons, each of which then relays to its own local procs. The crossover points between the algos are adjustable via MCA param, or you can simply demand that a specific algo be used.

3. orteds no longer exhibit two types of behavior: bootproxy or VM. Orteds now always behave like they are part of a virtual machine - they simply launch a job if mpirun tells them to do so. This is another step towards creating an "orteboot" functionality, but also provided a clean system for supporting message relaying.

Note one major impact of this commit: multiple daemons on a node cannot be supported any longer! Only a single daemon/node is now allowed.

This commit is known to break support for the following environments: POE, Xgrid, Xcpu, Windows. It has been tested on rsh, SLURM, and Bproc. Modifications for TM support have been made but could not be verified due to machine problems at LANL. Modifications for SGE have been made but could not be verified. The developers for the non-verified environments will be separately notified along with suggestions on how to fix the problems.

This commit was SVN r15007.
2007-06-12 13:28:54 +00:00

580 строки
18 KiB
C

/*
* Copyright (c) 2004-2005 The Trustees of Indiana University and Indiana
* University Research and Technology
* Corporation. All rights reserved.
* Copyright (c) 2004-2006 The University of Tennessee and The University
* of Tennessee Research Foundation. All rights
* reserved.
* Copyright (c) 2004-2005 High Performance Computing Center Stuttgart,
* University of Stuttgart. All rights reserved.
* Copyright (c) 2004-2005 The Regents of the University of California.
* All rights reserved.
* Copyright (c) 2006 Cisco Systems, Inc. All rights reserved.
* Copyright (c) 2007 Los Alamos National Security, LLC. All rights
* reserved.
* $COPYRIGHT$
*
* Additional copyrights may follow
*
* $HEADER$
*
* These symbols are in a file by themselves to provide nice linker
* semantics. Since linkers generally pull in symbols by object
* files, keeping these symbols as the only symbols in this file
* prevents utility programs such as "ompi_info" from having to import
* entire components just to query their version and parameters.
*/
#include "orte_config.h"
#include "orte/orte_constants.h"
#if HAVE_UNISTD_H
#include <unistd.h>
#endif
#include <signal.h>
#ifdef HAVE_SYS_TYPES_H
#include <sys/types.h>
#endif
#ifdef HAVE_SYS_WAIT_H
#include <sys/wait.h>
#endif
#ifdef HAVE_SCHED_H
#include <sched.h>
#endif
#ifdef HAVE_SYS_TIME_H
#include <sys/time.h>
#endif
#include <errno.h>
#include <tm.h>
#include "opal/mca/installdirs/installdirs.h"
#include "opal/threads/condition.h"
#include "opal/event/event.h"
#include "opal/util/argv.h"
#include "opal/util/output.h"
#include "opal/util/opal_environ.h"
#include "opal/util/show_help.h"
#include "opal/util/path.h"
#include "opal/util/basename.h"
#include "opal/mca/base/mca_base_param.h"
#include "opal/runtime/opal_progress.h"
#include "orte/orte_types.h"
#include "orte/runtime/runtime.h"
#include "orte/runtime/orte_wait.h"
#include "orte/runtime/orte_wakeup.h"
#include "orte/mca/pls/pls.h"
#include "orte/mca/errmgr/errmgr.h"
#include "orte/mca/smr/smr.h"
#include "orte/mca/gpr/gpr.h"
#include "orte/mca/sds/base/base.h"
#include "orte/mca/rmaps/rmaps.h"
#include "orte/mca/rml/rml.h"
#include "orte/mca/ns/ns.h"
#include "orte/mca/pls/base/base.h"
#include "orte/mca/pls/base/pls_private.h"
#include "pls_tm.h"
/*
* Local functions
*/
static int pls_tm_launch_job(orte_jobid_t jobid);
static int pls_tm_terminate_job(orte_jobid_t jobid, struct timeval *timeout, opal_list_t *attrs);
static int pls_tm_terminate_orteds(struct timeval *timeout, opal_list_t *attrs);
static int pls_tm_terminate_proc(const orte_process_name_t *name);
static int pls_tm_signal_job(orte_jobid_t jobid, int32_t signal, opal_list_t *attrs);
static int pls_tm_signal_proc(const orte_process_name_t *name, int32_t signal);
static int pls_tm_finalize(void);
static int pls_tm_connect(void);
static int pls_tm_disconnect(void);
/*
* Local variables
*/
/*
* Global variable
*/
orte_pls_base_module_t orte_pls_tm_module = {
pls_tm_launch_job,
pls_tm_terminate_job,
pls_tm_terminate_orteds,
pls_tm_terminate_proc,
pls_tm_signal_job,
pls_tm_signal_proc,
pls_tm_finalize
};
/* When working in this function, ALWAYS jump to "cleanup" if
* you encounter an error so that orterun will be woken up and
* the job can cleanly terminate
*/
static int pls_tm_launch_job(orte_jobid_t jobid)
{
orte_job_map_t *map = NULL;
opal_list_item_t *item;
size_t num_nodes;
orte_vpid_t vpid;
int node_name_index;
int proc_name_index;
char *param;
char **env = NULL;
char *var;
char **argv = NULL;
int argc;
int rc;
bool connected = false;
uint launched = 0, i;
char *bin_base = NULL, *lib_base = NULL;
tm_event_t *tm_events = NULL;
tm_task_id *tm_task_ids = NULL;
int local_err;
tm_event_t event;
struct timeval launchstart, launchstop, completionstart, completionstop;
struct timeval jobstart, jobstop;
int maxtime=0, mintime=99999999, maxiter = 0, miniter = 0, deltat;
float avgtime=0.0;
bool failed_launch = true;
/* check for timing request - get start time if so */
if (mca_pls_tm_component.timing) {
if (0 != gettimeofday(&jobstart, NULL)) {
opal_output(0, "pls_tm: could not obtain job start time");
}
}
/* Query the map for this job.
* We need the entire mapping for a couple of reasons:
* - need the prefix to start with.
* - need to know if we are launching on a subset of the allocated nodes
*/
rc = orte_rmaps.get_job_map(&map, jobid);
if (ORTE_SUCCESS != rc) {
ORTE_ERROR_LOG(rc);
goto cleanup;
}
/* Iterate through each of the nodes and check to see if we have
* a valid launch_id (must be > 0). If not, then error out as
* we cannot do anything
*/
for (item = opal_list_get_first(&map->nodes);
item != opal_list_get_end(&map->nodes);
item = opal_list_get_next(item)) {
orte_mapped_node_t* node = (orte_mapped_node_t*)item;
if (node->launch_id < 0) {
opal_show_help("help-pls-tm.txt", "tm-bad-launchid",
true, node->nodename, node->launch_id);
goto cleanup;
}
}
/* account for any reuse of daemons */
if (ORTE_SUCCESS != (rc = orte_pls_base_launch_on_existing_daemons(map))) {
ORTE_ERROR_LOG(rc);
goto cleanup;
}
num_nodes = map->num_new_daemons;
if (0 == num_nodes) {
/* must have been launched on existing daemons - just return */
failed_launch = false;
rc = ORTE_SUCCESS;
goto cleanup;
}
/* Allocate a bunch of TM events to use for tm_spawn()ing */
tm_events = malloc(sizeof(tm_event_t) * num_nodes);
if (NULL == tm_events) {
rc = ORTE_ERR_OUT_OF_RESOURCE;
ORTE_ERROR_LOG(rc);
goto cleanup;
}
tm_task_ids = malloc(sizeof(tm_task_id) * num_nodes);
if (NULL == tm_task_ids) {
rc = ORTE_ERR_OUT_OF_RESOURCE;
ORTE_ERROR_LOG(rc);
goto cleanup;
}
/* add the daemon command (as specified by user) */
argv = opal_argv_split(mca_pls_tm_component.orted, ' ');
argc = opal_argv_count(argv);
opal_argv_append(&argc, &argv, "--no-daemonize");
/* Add basic orted command line options */
orte_pls_base_orted_append_basic_args(&argc, &argv,
&proc_name_index,
&node_name_index,
(vpid + num_nodes));
if (mca_pls_tm_component.debug) {
param = opal_argv_join(argv, ' ');
if (NULL != param) {
opal_output(0, "pls:tm: final top-level argv:");
opal_output(0, "pls:tm: %s", param);
free(param);
}
}
rc = pls_tm_connect();
if (ORTE_SUCCESS != rc) {
goto cleanup;
}
connected = true;
/* Figure out the basenames for the libdir and bindir. There is a
lengthy comment about this in pls_rsh_module.c explaining all
the rationale for how / why we're doing this. */
lib_base = opal_basename(opal_install_dirs.libdir);
bin_base = opal_basename(opal_install_dirs.bindir);
/* setup environment */
env = opal_argv_copy(environ);
var = mca_base_param_environ_variable("seed",NULL,NULL);
opal_setenv(var, "0", true, &env);
/* clean out any MCA component selection directives that
* won't work on remote nodes
*/
orte_pls_base_purge_mca_params(&env);
/* If we have a prefix, then modify the PATH and
LD_LIBRARY_PATH environment variables. We only allow
a single prefix to be specified. Since there will
always be at least one app_context, we take it from
there
*/
if (NULL != map->apps[0]->prefix_dir) {
char *newenv;
for (i = 0; NULL != env && NULL != env[i]; ++i) {
/* Reset PATH */
if (0 == strncmp("PATH=", env[i], 5)) {
asprintf(&newenv, "%s/%s:%s",
map->apps[0]->prefix_dir, bin_base, env[i] + 5);
if (mca_pls_tm_component.debug) {
opal_output(0, "pls:tm: resetting PATH: %s",
newenv);
}
opal_setenv("PATH", newenv, true, &env);
free(newenv);
}
/* Reset LD_LIBRARY_PATH */
else if (0 == strncmp("LD_LIBRARY_PATH=", env[i], 16)) {
asprintf(&newenv, "%s/%s:%s",
map->apps[0]->prefix_dir, lib_base, env[i] + 16);
if (mca_pls_tm_component.debug) {
opal_output(0, "pls:tm: resetting LD_LIBRARY_PATH: %s",
newenv);
}
opal_setenv("LD_LIBRARY_PATH", newenv, true, &env);
free(newenv);
}
}
}
/* Iterate through each of the nodes and spin
* up a daemon.
*/
for (item = opal_list_get_first(&map->nodes);
item != opal_list_get_end(&map->nodes);
item = opal_list_get_next(item)) {
orte_mapped_node_t* node = (orte_mapped_node_t*)item;
orte_process_name_t* name;
char* name_string;
/* setup node name */
free(argv[node_name_index]);
argv[node_name_index] = strdup(node->nodename);
/* initialize daemons process name */
rc = orte_ns.create_process_name(&name, node->cell, 0, vpid);
if (ORTE_SUCCESS != rc) {
ORTE_ERROR_LOG(rc);
goto cleanup;
}
/* setup per-node options */
if (mca_pls_tm_component.debug ||
mca_pls_tm_component.verbose) {
opal_output(0, "pls:tm: launching on node %s",
node->nodename);
}
/* setup process name */
rc = orte_ns.get_proc_name_string(&name_string, name);
if (ORTE_SUCCESS != rc) {
opal_output(0, "pls:tm: unable to create process name");
goto cleanup;
}
free(argv[proc_name_index]);
argv[proc_name_index] = strdup(name_string);
/* exec the daemon */
if (mca_pls_tm_component.debug) {
param = opal_argv_join(argv, ' ');
if (NULL != param) {
opal_output(0, "pls:tm: executing: %s", param);
free(param);
}
}
/* check for timing request - get start time if so */
if (mca_pls_tm_component.timing) {
if (0 != gettimeofday(&launchstart, NULL)) {
opal_output(0, "pls_tm: could not obtain start time");
launchstart.tv_sec = 0;
launchstart.tv_usec = 0;
}
}
rc = tm_spawn(argc, argv, env, node->launch_id, tm_task_ids + launched, tm_events + launched);
if (TM_SUCCESS != rc) {
opal_show_help("help-pls-tm.txt", "tm-spawn-failed",
true, argv[0], node->nodename, node->launch_id);
rc = ORTE_ERROR;
goto cleanup;
}
/* check for timing request - get stop time and process if so */
if (mca_pls_tm_component.timing) {
if (0 != gettimeofday(&launchstop, NULL)) {
opal_output(0, "pls_tm: could not obtain stop time");
} else {
deltat = (launchstop.tv_sec - launchstart.tv_sec)*1000000 +
(launchstop.tv_usec - launchstart.tv_usec);
avgtime = avgtime + deltat / num_nodes;
if (deltat < mintime) {
mintime = deltat;
miniter = launched;
}
if (deltat > maxtime) {
maxtime = deltat;
maxiter = launched;
}
}
}
launched++;
++vpid;
free(name);
/* Allow some progress to occur */
opal_event_loop(OPAL_EVLOOP_NONBLOCK);
}
if (mca_pls_tm_component.debug) {
opal_output(0, "pls:tm:launch: finished spawning orteds\n");
}
/* check for timing request - get start time for launch completion */
if (mca_pls_tm_component.timing) {
if (0 != gettimeofday(&completionstart, NULL)) {
opal_output(0, "pls_tm: could not obtain completion start time");
completionstart.tv_sec = 0;
completionstart.tv_usec = 0;
}
}
/* TM poll for all the spawns */
for (i = 0; i < launched; ++i) {
rc = tm_poll(TM_NULL_EVENT, &event, 1, &local_err);
if (TM_SUCCESS != rc) {
errno = local_err;
opal_output(0, "pls:tm: failed to poll for a spawned proc, return status = %d", rc);
goto cleanup;
}
}
/* if we get here, then everything launched okay - record that fact */
failed_launch = false;
/* check for timing request - get stop time for launch completion and report */
if (mca_pls_tm_component.timing) {
if (0 != gettimeofday(&completionstop, NULL)) {
opal_output(0, "pls_tm: could not obtain completion stop time");
} else {
deltat = (launchstop.tv_sec - launchstart.tv_sec)*1000000 +
(launchstop.tv_usec - launchstart.tv_usec);
opal_output(0, "pls_tm: launch completion required %d usec", deltat);
}
opal_output(0, "pls_tm: Launch statistics:");
opal_output(0, "pls_tm: Average time to launch an orted: %f usec", avgtime);
opal_output(0, "pls_tm: Max time to launch an orted: %d usec at iter %d", maxtime, maxiter);
opal_output(0, "pls_tm: Min time to launch an orted: %d usec at iter %d", mintime, miniter);
}
cleanup:
if (NULL != map) {
OBJ_RELEASE(map);
}
if (NULL != argv) {
opal_argv_free(argv);
}
if (NULL != env) {
opal_argv_free(env);
}
if (connected) {
pls_tm_disconnect();
}
if (NULL != tm_events) {
free(tm_events);
}
if (NULL != tm_task_ids) {
free(tm_task_ids);
}
if (NULL != lib_base) {
free(lib_base);
}
if (NULL != bin_base) {
free(bin_base);
}
/* check for failed launch - if so, force terminate */
if (failed_launch) {
if (ORTE_SUCCESS != (rc = orte_smr.set_job_state(jobid, ORTE_JOB_STATE_FAILED_TO_START))) {
ORTE_ERROR_LOG(rc);
}
if (ORTE_SUCCESS != (rc = orte_wakeup(jobid))) {
ORTE_ERROR_LOG(rc);
}
}
/* check for timing request - get stop time and process if so */
if (mca_pls_tm_component.timing) {
if (0 != gettimeofday(&jobstop, NULL)) {
opal_output(0, "pls_tm: could not obtain stop time");
} else {
deltat = (jobstop.tv_sec - jobstart.tv_sec)*1000000 +
(jobstop.tv_usec - jobstart.tv_usec);
opal_output(0, "pls_tm: launch of entire job required %d usec", deltat);
}
}
if (mca_pls_tm_component.debug) {
opal_output(0, "pls:tm:launch: finished\n");
}
return rc;
}
static int pls_tm_terminate_job(orte_jobid_t jobid, struct timeval *timeout, opal_list_t *attrs)
{
int rc;
/* order all of the daemons to kill their local procs for this job */
if (ORTE_SUCCESS != (rc = orte_pls_base_orted_kill_local_procs(jobid, timeout, attrs))) {
ORTE_ERROR_LOG(rc);
}
return rc;
}
/**
* Terminate the orteds for a given job
*/
int pls_tm_terminate_orteds(struct timeval *timeout, opal_list_t *attrs)
{
int rc;
/* now tell them to die! */
if (ORTE_SUCCESS != (rc = orte_pls_base_orted_exit(timeout, attrs))) {
ORTE_ERROR_LOG(rc);
}
return rc;
}
/*
* TM can't kill individual processes -- PBS will kill the entire job
*/
static int pls_tm_terminate_proc(const orte_process_name_t *name)
{
if (mca_pls_tm_component.debug) {
opal_output(0, "pls:tm:terminate_proc: not supported");
}
return ORTE_ERR_NOT_SUPPORTED;
}
static int pls_tm_signal_job(orte_jobid_t jobid, int32_t signal, opal_list_t *attrs)
{
int rc;
/* order them to pass this signal to their local procs */
if (ORTE_SUCCESS != (rc = orte_pls_base_orted_signal_local_procs(jobid, signal, attrs))) {
ORTE_ERROR_LOG(rc);
}
return rc;
}
static int pls_tm_signal_proc(const orte_process_name_t *name, int32_t signal)
{
return ORTE_ERR_NOT_IMPLEMENTED;
}
/*
* Free stuff
*/
static int pls_tm_finalize(void)
{
int rc;
/* cleanup any pending recvs */
if (ORTE_SUCCESS != (rc = orte_pls_base_comm_stop())) {
ORTE_ERROR_LOG(rc);
}
return ORTE_SUCCESS;
}
static int pls_tm_connect(void)
{
int ret;
struct tm_roots tm_root;
int count, progress;
/* try a couple times to connect - might get busy signals every
now and then */
for (count = 0 ; count < 10; ++count) {
ret = tm_init(NULL, &tm_root);
if (TM_SUCCESS == ret) {
return ORTE_SUCCESS;
}
for (progress = 0 ; progress < 10 ; ++progress) {
opal_progress();
#if HAVE_SCHED_YIELD
sched_yield();
#endif
}
}
return ORTE_ERR_RESOURCE_BUSY;
}
static int pls_tm_disconnect(void)
{
tm_finalize();
return ORTE_SUCCESS;
}