diff --git a/src/slurmctld/job_mgr.c b/src/slurmctld/job_mgr.c
index b970700bf12fba4cbbe90e64c48214b1e13b80d8..628a87f6ef24ee89a7c1177f82a451d59ce95ce5 100644
--- a/src/slurmctld/job_mgr.c
+++ b/src/slurmctld/job_mgr.c
@@ -1440,9 +1440,8 @@ extern int job_allocate(job_desc_msg_t * job_specs, int immediate, int will_run,
no_alloc = test_only || too_fragmented ||
(!top_prio) || (!independent);
-
error_code = select_nodes(job_ptr, no_alloc);
-
+
if ((error_code == ESLURM_NODES_BUSY)
|| (error_code == ESLURM_JOB_HELD)
|| (error_code == ESLURM_REQUESTED_PART_CONFIG_UNAVAILABLE)) {
diff --git a/src/slurmctld/node_mgr.c b/src/slurmctld/node_mgr.c
index 209520ad8452e9959186776c0cb7a75381e4f311..f876860c6164cc49e6d4b9173865908cbd5d8dfb 100644
--- a/src/slurmctld/node_mgr.c
+++ b/src/slurmctld/node_mgr.c
@@ -20,7 +20,7 @@
* any later version.
*
* In addition, as a special exception, the copyright holders give permission
- * to link the code of portions of this program with the OpenSSL library under
+ * to link the code of portions of this program with the OpenSSL library under
* certain conditions as described in each individual source file, and
* distribute linked combinations including the two. You must obey the GNU
* General Public License in all respects for all of the code used other than
@@ -30,17 +30,6 @@
* version. If you delete this exception statement from all source files in
* the program, then also delete it here.
*
- * In addition, as a special exception, the copyright holders give permission
- * to link the code of portions of this program with the OpenSSL library under
- * certain conditions as described in each individual source file, and
- * distribute linked combinations including the two. You must obey the GNU
- * General Public License in all respects for all of the code used other than
- * OpenSSL. If you modify file(s) with this exception, you may extend this
- * exception to your version of the file(s), but you are not obligated to do
- * so. If you do not wish to do so, delete this exception statement from your
- * version. If you delete this exception statement from all source files in
- * the program, then also delete it here.
- *
* SLURM is distributed in the hope that it will be useful, but WITHOUT ANY
* WARRANTY; without even the implied warranty of MERCHANTABILITY or FITNESS
* FOR A PARTICULAR PURPOSE. See the GNU General Public License for more
diff --git a/src/slurmctld/node_scheduler.c b/src/slurmctld/node_scheduler.c
index 2846e4cefacc87acd7f9f5b8e30a014cfa39d68a..0c97fe1cc6ccf32f5224e471236cb8b2b4c309c4 100644
--- a/src/slurmctld/node_scheduler.c
+++ b/src/slurmctld/node_scheduler.c
@@ -293,10 +293,10 @@ _pick_best_load(struct job_record *job_ptr, bitstr_t * bitmap,
{
bitstr_t *no_load_bit, *light_load_bit, *heavy_load_bit;
int error_code;
-
+
_node_load_bitmaps(bitmap, &no_load_bit, &light_load_bit,
&heavy_load_bit);
-
+
/* first try to use idle nodes */
bit_and(bitmap, no_load_bit);
FREE_NULL_BITMAP(no_load_bit);
@@ -462,7 +462,7 @@ _pick_best_nodes(struct node_set *node_set_ptr, int node_set_size,
bool runable_avail = false; /* Job can run with available nodes */
int cr_enabled = 0;
int shared = 0;
-
+
if (node_set_size == 0) {
info("_pick_best_nodes: empty node set for selection");
return ESLURM_REQUESTED_NODE_CONFIG_UNAVAILABLE;
@@ -580,12 +580,18 @@ _pick_best_nodes(struct node_set *node_set_ptr, int node_set_size,
if (node_set_ptr[i].feature < min_feature)
min_feature = node_set_ptr[i].feature;
}
-
+
for (j = min_feature; j <= max_feature; j++) {
+ /* we use this var to go straight down the list if the
+ * first one doesn't work we go to the next until the
+ * list is empty.
+ */
+ int tries = 0;
for (i = 0; i < node_set_size; i++) {
bool pick_light_load = false;
if (node_set_ptr[i].feature != j)
continue;
+
if (!runable_ever) {
int cr_disabled = 0;
error_code = _add_node_set_info(
@@ -603,6 +609,7 @@ _pick_best_nodes(struct node_set *node_set_ptr, int node_set_size,
}
}
bit_and(node_set_ptr[i].my_bitmap, avail_node_bitmap);
+
if (cr_enabled) {
bit_and(node_set_ptr[i].my_bitmap,
partially_idle_node_bitmap);
@@ -633,7 +640,7 @@ _pick_best_nodes(struct node_set *node_set_ptr, int node_set_size,
}
node_set_ptr[i].nodes =
bit_set_count(node_set_ptr[i].my_bitmap);
- error_code = _add_node_set_info(&node_set_ptr[i],
+ error_code = _add_node_set_info(&node_set_ptr[i],
&avail_bitmap,
&avail_nodes,
&avail_cpus,
@@ -672,6 +679,7 @@ _pick_best_nodes(struct node_set *node_set_ptr, int node_set_size,
req_nodes,
false);
}
+
if (pick_code == SLURM_SUCCESS) {
if (bit_set_count(avail_bitmap) > max_nodes) {
/* end of tests for this feature */
@@ -682,8 +690,17 @@ _pick_best_nodes(struct node_set *node_set_ptr, int node_set_size,
if (cr_enabled)
FREE_NULL_BITMAP(
partially_idle_node_bitmap);
+
*select_bitmap = avail_bitmap;
return SLURM_SUCCESS;
+ } else {
+ /* reset the counters and start from the
+ * next node in the list */
+ FREE_NULL_BITMAP(avail_bitmap);
+ avail_nodes = 0;
+ avail_cpus = 0;
+ tries++;
+ i = tries;
}
}
@@ -941,8 +958,9 @@ extern int select_nodes(struct job_record *job_ptr, bool test_only)
error_code = ESLURM_REQUESTED_PART_CONFIG_UNAVAILABLE;
} else {
error_code = _pick_best_nodes(node_set_ptr, node_set_size,
- &select_bitmap, job_ptr, part_ptr,
- min_nodes, max_nodes, req_nodes);
+ &select_bitmap, job_ptr,
+ part_ptr, min_nodes, max_nodes,
+ req_nodes);
}
if (error_code) {
diff --git a/src/slurmctld/proc_req.c b/src/slurmctld/proc_req.c
index 7da3508678a5fa23321be3f2c1b682dad0d38ea2..b7dc4e168a695bdb1c0f7485804547867049982c 100644
--- a/src/slurmctld/proc_req.c
+++ b/src/slurmctld/proc_req.c
@@ -17,7 +17,7 @@
* any later version.
*
* In addition, as a special exception, the copyright holders give permission
- * to link the code of portions of this program with the OpenSSL library under
+ * to link the code of portions of this program with the OpenSSL library under
* certain conditions as described in each individual source file, and
* distribute linked combinations including the two. You must obey the GNU
* General Public License in all respects for all of the code used other than
@@ -104,7 +104,7 @@ inline static void _slurm_rpc_job_will_run(slurm_msg_t * msg);
inline static void _slurm_rpc_node_registration(slurm_msg_t * msg);
inline static void _slurm_rpc_node_select_info(slurm_msg_t * msg);
inline static void _slurm_rpc_job_alloc_info(slurm_msg_t * msg);
-inline static void _slurm_rpc_job_alloc_info_lite(slurm_msg_t * msg);
+inline static void _slurm_rpc_job_alloc_info_lite(slurm_msg_t * msg);
inline static void _slurm_rpc_ping(slurm_msg_t * msg);
inline static void _slurm_rpc_reconfigure_controller(slurm_msg_t * msg);
inline static void _slurm_rpc_requeue(slurm_msg_t * msg);