1dc12046f1
Adds several new mpirun options: * -bysocket - assign ranks on a node by socket. Effectively load balances the procs assigned to a node across the available sockets. Note that ranks can still be bound to a specific core within the socket, or to the entire socket - the mapping is independent of the binding. * -bind-to-socket - bind each rank to all the cores on the socket to which they are assigned. * -bind-to-core - currently the default behavior (maintained from prior default) * -npersocket N - launch N procs for every socket on a node. Note that this implies we know how many sockets are on a node. Mpirun will determine its local values. These can be overridden by provided values, either via MCA param or in a hostfile Similar features/options are provided at the board level for multi-board nodes. Documentation to follow... This commit was SVN r21791.
545 строки
20 KiB
C
545 строки
20 KiB
C
/*
|
|
* Copyright (c) 2004-2007 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.
|
|
* $COPYRIGHT$
|
|
*
|
|
* Additional copyrights may follow
|
|
*
|
|
* $HEADER$
|
|
*/
|
|
|
|
#include "orte_config.h"
|
|
#include "orte/constants.h"
|
|
#include "orte/types.h"
|
|
|
|
#include <errno.h>
|
|
#ifdef HAVE_UNISTD_H
|
|
#include <unistd.h>
|
|
#endif /* HAVE_UNISTD_H */
|
|
#ifdef HAVE_STRING_H
|
|
#include <string.h>
|
|
#endif /* HAVE_STRING_H */
|
|
|
|
#include "opal/mca/base/mca_base_param.h"
|
|
#include "opal/util/trace.h"
|
|
#include "opal/mca/carto/base/base.h"
|
|
|
|
#include "orte/util/show_help.h"
|
|
#include "orte/mca/errmgr/errmgr.h"
|
|
|
|
#include "orte/mca/rmaps/base/rmaps_private.h"
|
|
#include "orte/mca/rmaps/base/base.h"
|
|
#include "rmaps_topo.h"
|
|
|
|
static int topo_map(orte_job_t *jdata);
|
|
|
|
orte_rmaps_base_module_t orte_rmaps_topo_module = {
|
|
topo_map
|
|
};
|
|
|
|
/*
|
|
* Local variable
|
|
*/
|
|
static opal_list_item_t *cur_node_item = NULL;
|
|
static int ppn = 0;
|
|
|
|
/*
|
|
* Create a default mapping for the application, scheduling round
|
|
* robin by node.
|
|
*/
|
|
static int map_app_by_node(
|
|
orte_app_context_t* app,
|
|
orte_job_t* jdata,
|
|
orte_vpid_t vpid_start,
|
|
opal_list_t* nodes)
|
|
{
|
|
int rc = ORTE_SUCCESS;
|
|
opal_list_item_t *next;
|
|
orte_node_t *node;
|
|
orte_std_cntr_t num_alloc=0;
|
|
|
|
OPAL_TRACE(2);
|
|
|
|
/* This loop continues until all procs have been mapped or we run
|
|
out of resources. We determine that we have "run out of
|
|
resources" when all nodes have slots_max processes mapped to them,
|
|
thus there are no free slots for a process to be mapped, or we have
|
|
hit the soft limit on all nodes and are in a "no oversubscribe" state.
|
|
If we still have processes that haven't been mapped yet, then it's an
|
|
"out of resources" error.
|
|
|
|
In this scenario, we rely on the claim_slot function to handle the
|
|
oversubscribed case. The claim_slot function will leave a node on the
|
|
list until it either reaches slots_max OR reaches the
|
|
soft limit and the "no_oversubscribe" flag has been set - at which point,
|
|
the node will be removed to prevent any more processes from being mapped to
|
|
it. Since we are taking one slot from each node as we cycle through, the
|
|
list, oversubscription is automatically taken care of via this logic.
|
|
*/
|
|
|
|
while (num_alloc < app->num_procs) {
|
|
|
|
/** see if any nodes remain unused and available. We need to do this check
|
|
* each time since we may remove nodes from the list (as they become fully
|
|
* used) as we cycle through the loop */
|
|
if(0 >= opal_list_get_size(nodes) ) {
|
|
/* No more nodes to allocate :( */
|
|
orte_show_help("help-orte-rmaps-topo.txt", "orte-rmaps-topo:alloc-error",
|
|
true, app->num_procs, app->app);
|
|
return ORTE_ERR_SILENT;
|
|
}
|
|
|
|
/* Save the next node we can use before claiming slots, since
|
|
* we may need to prune the nodes list removing overused nodes.
|
|
* Wrap around to beginning if we are at the end of the list */
|
|
if (opal_list_get_end(nodes) == opal_list_get_next(cur_node_item)) {
|
|
next = opal_list_get_first(nodes);
|
|
}
|
|
else {
|
|
next = opal_list_get_next(cur_node_item);
|
|
}
|
|
|
|
/* Allocate a slot on this node */
|
|
node = (orte_node_t*) cur_node_item;
|
|
if (ORTE_SUCCESS != (rc = orte_rmaps_base_claim_slot(jdata, node, 1, app->idx,
|
|
nodes, jdata->map->oversubscribe, true, NULL))) {
|
|
/** if the code is ORTE_ERR_NODE_FULLY_USED, then we know this
|
|
* really isn't an error - we just need to break from the loop
|
|
* since the node is fully used up. For now, just don't report
|
|
* an error
|
|
*/
|
|
if (ORTE_ERR_NODE_FULLY_USED != rc) {
|
|
ORTE_ERROR_LOG(rc);
|
|
return rc;
|
|
}
|
|
}
|
|
|
|
++num_alloc;
|
|
|
|
cur_node_item = next;
|
|
}
|
|
|
|
return ORTE_SUCCESS;
|
|
}
|
|
|
|
|
|
/*
|
|
* Create a default mapping for the application, scheduling one round
|
|
* robin by slot.
|
|
*/
|
|
static int map_app_by_slot(
|
|
orte_app_context_t* app,
|
|
orte_job_t* jdata,
|
|
orte_vpid_t vpid_start,
|
|
opal_list_t* nodes)
|
|
{
|
|
int rc = ORTE_SUCCESS;
|
|
orte_std_cntr_t i, num_slots_to_take;
|
|
orte_node_t *node;
|
|
opal_list_item_t *next;
|
|
orte_std_cntr_t num_alloc=0;
|
|
|
|
OPAL_TRACE(2);
|
|
|
|
/* This loop continues until all procs have been mapped or we run
|
|
out of resources. We determine that we have "run out of
|
|
resources" when either all nodes have slots_max processes mapped to them,
|
|
(thus there are no free slots for a process to be mapped), OR all nodes
|
|
have reached their soft limit and the user directed us to "no oversubscribe".
|
|
If we still have processes that haven't been mapped yet, then it's an
|
|
"out of resources" error. */
|
|
|
|
while ( num_alloc < app->num_procs) {
|
|
|
|
/** see if any nodes remain unused and available. We need to do this check
|
|
* each time since we may remove nodes from the list (as they become fully
|
|
* used) as we cycle through the loop */
|
|
if(0 >= opal_list_get_size(nodes) ) {
|
|
/* Everything is at max usage! :( */
|
|
orte_show_help("help-orte-rmaps-topo.txt", "orte-rmaps-topo:alloc-error",
|
|
true, app->num_procs, app->app);
|
|
return ORTE_ERR_SILENT;
|
|
}
|
|
|
|
/* Save the next node we can use before claiming slots, since
|
|
* we may need to prune the nodes list removing overused nodes.
|
|
* Wrap around to beginning if we are at the end of the list */
|
|
if (opal_list_get_end(nodes) == opal_list_get_next(cur_node_item)) {
|
|
next = opal_list_get_first(nodes);
|
|
}
|
|
else {
|
|
next = opal_list_get_next(cur_node_item);
|
|
}
|
|
|
|
/** declare a shorter name for convenience in the code below */
|
|
node = (orte_node_t*) cur_node_item;
|
|
|
|
/* If we have available slots on this node, claim all of them
|
|
* If node_slots == 0, assume 1 slot for that node.
|
|
* JJH - is this assumption fully justified?
|
|
*
|
|
* If we are now oversubscribing the nodes, then we still take:
|
|
* (a) if the node has not been used yet, we take a full node_slots
|
|
* (b) if some of the slots are in-use, then we take the number of
|
|
* remaining slots before hitting the soft limit (node_slots)
|
|
* (c) if we are at or above the soft limit, we take a full node_slots
|
|
*
|
|
* Note: if node_slots is zero, then we always just take 1 slot
|
|
*
|
|
* We continue this process until either everything is done,
|
|
* or all nodes have hit their hard limit. This algorithm ensures we
|
|
* fully utilize each node before oversubscribing, and preserves the ratio
|
|
* of processes between the nodes thereafter (e.g., if one node has twice as
|
|
* many processes as another before oversubscribing, it will continue
|
|
* to do so after oversubscribing).
|
|
*/
|
|
if (0 == node->slots_inuse ||
|
|
node->slots_inuse >= node->slots) {
|
|
num_slots_to_take = (node->slots == 0) ? 1 : node->slots;
|
|
} else {
|
|
num_slots_to_take = node->slots - node->slots_inuse;
|
|
}
|
|
|
|
/* check if we are in npernode mode - if so, then set the num_slots_to_take
|
|
* to the num_per_node
|
|
*/
|
|
if (0 < jdata->map->npernode) {
|
|
num_slots_to_take = jdata->map->npernode;
|
|
}
|
|
|
|
for( i = 0; i < num_slots_to_take; ++i) {
|
|
if (ORTE_SUCCESS != (rc = orte_rmaps_base_claim_slot(jdata, node, 1, app->idx,
|
|
nodes, jdata->map->oversubscribe, true, NULL))) {
|
|
/** if the code is ORTE_ERR_NODE_FULLY_USED, then we know this
|
|
* really isn't an error - we just need to break from the loop
|
|
* since the node is fully used up. For now, just don't report
|
|
* an error
|
|
*/
|
|
if (ORTE_ERR_NODE_FULLY_USED != rc) {
|
|
ORTE_ERROR_LOG(rc);
|
|
return rc;
|
|
}
|
|
}
|
|
|
|
/* Update the number of procs allocated */
|
|
++num_alloc;
|
|
|
|
/** if all the procs have been mapped, we return */
|
|
if (num_alloc == app->num_procs) {
|
|
return ORTE_SUCCESS;
|
|
}
|
|
|
|
/* if we have fully used up this node
|
|
* OR we are at our ppn and loadbalancing, then break from the loop
|
|
*/
|
|
if (ORTE_ERR_NODE_FULLY_USED == rc ||
|
|
(orte_rmaps_base.loadbalance && (int)node->num_procs >= ppn)) {
|
|
break;
|
|
}
|
|
}
|
|
|
|
/* we move on to the next node in all cases EXCEPT if we came
|
|
* out of the loop without having taken a full bite AND the
|
|
* node is NOT max'd out
|
|
*
|
|
*/
|
|
if (i < (num_slots_to_take-1) &&
|
|
ORTE_ERR_NODE_FULLY_USED != rc &&
|
|
(orte_rmaps_base.loadbalance && (int)node->num_procs < ppn)) {
|
|
continue;
|
|
}
|
|
cur_node_item = next;
|
|
}
|
|
|
|
return ORTE_SUCCESS;
|
|
}
|
|
|
|
|
|
/*
|
|
* Create a topo-aware mapping for the job.
|
|
*/
|
|
static int topo_map(orte_job_t *jdata)
|
|
{
|
|
orte_job_map_t *map;
|
|
orte_app_context_t *app, **apps;
|
|
orte_std_cntr_t i;
|
|
opal_list_t node_list;
|
|
opal_list_item_t *item;
|
|
orte_node_t *node, *nd1;
|
|
orte_vpid_t vpid_start;
|
|
orte_std_cntr_t num_nodes, num_slots;
|
|
int rc;
|
|
orte_std_cntr_t slots_per_node;
|
|
opal_carto_graph_t *graph;
|
|
opal_carto_base_node_t *crnode;
|
|
opal_value_array_t distance;
|
|
|
|
OPAL_TRACE(1);
|
|
|
|
/* conveniece def */
|
|
map = jdata->map;
|
|
apps = (orte_app_context_t**)jdata->apps->addr;
|
|
|
|
/* start at the beginning... */
|
|
vpid_start = 0;
|
|
jdata->num_procs = 0;
|
|
|
|
/* get the graph of nodes */
|
|
if (ORTE_SUCCESS != (rc = opal_carto_base_get_host_graph(&graph, "SLOT"))) {
|
|
ORTE_ERROR_LOG(rc);
|
|
return rc;
|
|
}
|
|
|
|
/* cycle through the app_contexts, mapping them sequentially */
|
|
for(i=0; i < jdata->num_apps; i++) {
|
|
app = apps[i];
|
|
|
|
/* if the number of processes wasn't specified, then we know there can be only
|
|
* one app_context allowed in the launch, and that we are to launch it across
|
|
* all available slots. We'll double-check the single app_context rule first
|
|
*/
|
|
if (0 == app->num_procs && 1 < jdata->num_apps) {
|
|
orte_show_help("help-orte-rmaps-rr.txt", "orte-rmaps-rr:multi-apps-and-zero-np",
|
|
true, jdata->num_apps, NULL);
|
|
rc = ORTE_ERR_SILENT;
|
|
goto error;
|
|
}
|
|
|
|
/* for each app_context, we have to get the list of nodes that it can
|
|
* use since that can now be modified with a hostfile and/or -host
|
|
* option
|
|
*/
|
|
OBJ_CONSTRUCT(&node_list, opal_list_t);
|
|
if(ORTE_SUCCESS != (rc = orte_rmaps_base_get_target_nodes(&node_list, &num_slots, app,
|
|
map->policy))) {
|
|
ORTE_ERROR_LOG(rc);
|
|
goto error;
|
|
}
|
|
num_nodes = (orte_std_cntr_t)opal_list_get_size(&node_list);
|
|
|
|
/* if a bookmark exists from some prior mapping, set us to start there */
|
|
if (NULL != jdata->bookmark) {
|
|
cur_node_item = NULL;
|
|
/* find this node on the list */
|
|
for (item = opal_list_get_first(&node_list);
|
|
item != opal_list_get_end(&node_list);
|
|
item = opal_list_get_next(item)) {
|
|
node = (orte_node_t*)item;
|
|
|
|
if (node->index == jdata->bookmark->index) {
|
|
cur_node_item = item;
|
|
break;
|
|
}
|
|
}
|
|
/* see if we found it - if not, just start at the beginning */
|
|
if (NULL == cur_node_item) {
|
|
cur_node_item = opal_list_get_first(&node_list);
|
|
}
|
|
} else {
|
|
/* if no bookmark, then just start at the beginning of the list */
|
|
cur_node_item = opal_list_get_first(&node_list);
|
|
}
|
|
|
|
/* order this list by network nearness - i.e., the next item in the
|
|
* list should be the node that is closest [in a network sense] to
|
|
* the prior item in the list
|
|
*
|
|
* RHC: start the list with the bookmark nodeas this is where
|
|
* we would start mapping
|
|
*/
|
|
node = (orte_node_t*)cur_node_item;
|
|
if (NULL == (crnode = opal_carto_base_find_node(graph, node->name))) {
|
|
ORTE_ERROR_LOG(ORTE_ERR_NOT_FOUND);
|
|
rc = ORTE_ERR_NOT_FOUND;
|
|
goto error;
|
|
}
|
|
OBJ_CONSTRUCT(&distance, opal_value_array_t);
|
|
if (ORTE_SUCCESS != (rc = opal_carto_base_get_nodes_distance(graph, crnode,
|
|
"SLOT", &distance))) {
|
|
ORTE_ERROR_LOG(rc);
|
|
goto error;
|
|
}
|
|
/* cycle through the nodes in the distance array - these
|
|
* should be in order based on distance
|
|
*/
|
|
#if 0
|
|
/* RHC: need to create a working list of nodes that is ordered
|
|
* according to distance. The get_nodes_distance function returns
|
|
* this, but it covers -all- nodes, so we have to filter that
|
|
* against the allocated node list to create the new
|
|
* working_node_list
|
|
*/
|
|
for (i=0; i < distance.size; i++) {
|
|
if
|
|
}
|
|
for (item = opal_list_get_first(&node_list);
|
|
item != opal_list_get_end(&node_list);
|
|
item = opal_list_get_next(item)) {
|
|
node = (orte_node_t*)item;
|
|
|
|
if (NULL == (crnode = opal_carto.find_node(graph, node->name))) {
|
|
ORTE_ERROR_LOG(ORTE_ERR_NOT_FOUND);
|
|
rc = ORTE_ERR_NOT_FOUND;
|
|
goto error;
|
|
}
|
|
|
|
/* look this node up in the distance array */
|
|
#endif
|
|
|
|
/* is this node oversubscribed? */
|
|
node = (orte_node_t*)cur_node_item;
|
|
if (node->slots_inuse > node->slots) {
|
|
/* work down the list - is there another node that
|
|
* would not be oversubscribed?
|
|
*/
|
|
if (cur_node_item != opal_list_get_end(&node_list)) {
|
|
item = opal_list_get_next(cur_node_item);
|
|
} else {
|
|
item = opal_list_get_first(&node_list);
|
|
}
|
|
while (item != cur_node_item) {
|
|
nd1 = (orte_node_t*)item;
|
|
if (nd1->slots_inuse < nd1->slots) {
|
|
/* this node is not oversubscribed! use it! */
|
|
cur_node_item = item;
|
|
goto proceed;
|
|
}
|
|
if (item == opal_list_get_end(&node_list)) {
|
|
item = opal_list_get_first(&node_list);
|
|
} else {
|
|
item= opal_list_get_next(item);
|
|
}
|
|
}
|
|
/* if we get here, then we cycled all the way around the
|
|
* list without finding a better answer - just use what
|
|
* we have
|
|
*/
|
|
}
|
|
|
|
proceed:
|
|
if (map->npernode == 1) {
|
|
/* there are three use-cases that we need to deal with:
|
|
* (a) if -np was not provided, then we just use the number of nodes
|
|
* (b) if -np was provided AND #procs > #nodes, then error out
|
|
* (c) if -np was provided AND #procs <= #nodes, then launch
|
|
* the specified #procs one/node. In this case, we just
|
|
* leave app->num_procs alone
|
|
*/
|
|
if (0 == app->num_procs) {
|
|
app->num_procs = num_nodes;
|
|
} else if (app->num_procs > num_nodes) {
|
|
orte_show_help("help-orte-rmaps-rr.txt", "orte-rmaps-rr:per-node-and-too-many-procs",
|
|
true, app->num_procs, num_nodes, NULL);
|
|
rc = ORTE_ERR_SILENT;
|
|
goto error;
|
|
}
|
|
} else if (map->npernode > 1) {
|
|
/* first, let's check to see if there are enough slots/node to
|
|
* meet the request - error out if not
|
|
*/
|
|
slots_per_node = num_slots / num_nodes;
|
|
if (map->npernode > slots_per_node) {
|
|
orte_show_help("help-orte-rmaps-rr.txt", "orte-rmaps-rr:n-per-node-and-not-enough-slots",
|
|
true, map->npernode, slots_per_node, NULL);
|
|
rc = ORTE_ERR_SILENT;
|
|
goto error;
|
|
}
|
|
/* there are three use-cases that we need to deal with:
|
|
* (a) if -np was not provided, then we just use the n/node * #nodes
|
|
* (b) if -np was provided AND #procs > (n/node * #nodes), then error out
|
|
* (c) if -np was provided AND #procs <= (n/node * #nodes), then launch
|
|
* the specified #procs n/node. In this case, we just
|
|
* leave app->num_procs alone
|
|
*/
|
|
if (0 == app->num_procs) {
|
|
/* set the num_procs to equal the specified num/node * the number of nodes */
|
|
app->num_procs = map->npernode * num_nodes;
|
|
} else if (app->num_procs > (map->npernode * num_nodes)) {
|
|
orte_show_help("help-orte-rmaps-rr.txt", "orte-rmaps-rr:n-per-node-and-too-many-procs",
|
|
true, app->num_procs, map->npernode, num_nodes, num_slots, NULL);
|
|
rc = ORTE_ERR_SILENT;
|
|
goto error;
|
|
}
|
|
} else if (0 == app->num_procs) {
|
|
/** set the num_procs to equal the number of slots on these mapped nodes - if
|
|
user has specified "-bynode", then set it to the number of nodes
|
|
*/
|
|
if (map->policy & ORTE_MAPPING_BYNODE) {
|
|
app->num_procs = num_nodes;
|
|
} else if (map->policy & ORTE_MAPPING_BYSLOT) {
|
|
app->num_procs = num_slots;
|
|
} else {
|
|
/* we can't handle this - it should have been set when we got
|
|
* the map info. If it wasn't, then we can only error out
|
|
*/
|
|
orte_show_help("help-orte-rmaps-rr.txt", "orte-rmaps-rr:no-np-and-user-map",
|
|
true, app->num_procs, map->npernode, num_nodes, num_slots, NULL);
|
|
rc = ORTE_ERR_SILENT;
|
|
goto error;
|
|
}
|
|
}
|
|
|
|
/** track the total number of processes we mapped */
|
|
jdata->num_procs += app->num_procs;
|
|
|
|
/* Make assignments */
|
|
if (map->policy == ORTE_MAPPING_BYNODE) {
|
|
rc = map_app_by_node(app, jdata, vpid_start, &node_list);
|
|
} else {
|
|
rc = map_app_by_slot(app, jdata, vpid_start, &node_list);
|
|
}
|
|
|
|
/* update the starting vpid for the next app_context */
|
|
vpid_start += app->num_procs;
|
|
|
|
if (ORTE_SUCCESS != rc) {
|
|
ORTE_ERROR_LOG(rc);
|
|
goto error;
|
|
}
|
|
|
|
/* save the bookmark */
|
|
jdata->bookmark = (orte_node_t*)cur_node_item;
|
|
|
|
/* cleanup the node list - it can differ from one app_context
|
|
* to another, so we have to get it every time
|
|
*/
|
|
while(NULL != (item = opal_list_remove_first(&node_list))) {
|
|
OBJ_RELEASE(item);
|
|
}
|
|
OBJ_DESTRUCT(&node_list);
|
|
}
|
|
|
|
/* compute and save convenience values */
|
|
if (ORTE_SUCCESS != (rc = orte_rmaps_base_compute_local_ranks(jdata))) {
|
|
ORTE_ERROR_LOG(rc);
|
|
return rc;
|
|
}
|
|
|
|
/* define the daemons that we will use for this job */
|
|
if (ORTE_SUCCESS != (rc = orte_rmaps_base_define_daemons(map))) {
|
|
ORTE_ERROR_LOG(rc);
|
|
return rc;
|
|
}
|
|
|
|
return ORTE_SUCCESS;
|
|
|
|
error:
|
|
while(NULL != (item = opal_list_remove_first(&node_list))) {
|
|
OBJ_RELEASE(item);
|
|
}
|
|
OBJ_DESTRUCT(&node_list);
|
|
|
|
return rc;
|
|
}
|
|
|
|
|