diff --git a/NEWS b/NEWS
index 567b063ad67b061f60991c167842ce3bdfe15308..326f121e5c25a30aaff10537efa72dad8ad89b51 100644
--- a/NEWS
+++ b/NEWS
@@ -8,6 +8,8 @@ documents those changes that are of interest to users and admins.
-- Fix several memory leaks.
-- In scontrol, show job's Requeue value, permit change of Requeue and COmment
values.
+ -- In slurmctld job record, add QOS (quality of service) value for accounting
+ purposes with Maui and Moab.
* Changes in SLURM 1.3.0-pre10
==============================
diff --git a/src/slurmctld/job_mgr.c b/src/slurmctld/job_mgr.c
index 775e94e4306789b233a1839469ba170cff7fe446..2a9b5fbe2b66f69570e4f926c4d6df0758a029a9 100644
--- a/src/slurmctld/job_mgr.c
+++ b/src/slurmctld/job_mgr.c
@@ -101,6 +101,8 @@ static int hash_table_size = 0;
static int job_count = 0; /* job's in the system */
static uint32_t job_id_sequence = 0; /* first job_id to assign new job */
static struct job_record **job_hash = NULL;
+static bool wiki_sched = false;
+static bool wiki_sched_test = false;
/* Local functions */
static void _add_job_hash(struct job_record *job_ptr);
@@ -487,6 +489,7 @@ static void _dump_job_state(struct job_record *dump_job_ptr, Buf buffer)
pack16(dump_job_ptr->kill_on_step_done, buffer);
pack16(dump_job_ptr->batch_flag, buffer);
pack16(dump_job_ptr->mail_type, buffer);
+ pack16(dump_job_ptr->qos, buffer);
pack16(dump_job_ptr->state_reason, buffer);
packstr(dump_job_ptr->resp_host, buffer);
@@ -542,7 +545,7 @@ static int _load_job_state(Buf buffer)
total_procs;
time_t start_time, end_time, suspend_time, pre_sus_time, tot_sus_time;
uint16_t job_state, next_step_id, details, batch_flag, step_flag;
- uint16_t kill_on_node_fail, kill_on_step_done;
+ uint16_t kill_on_node_fail, kill_on_step_done, qos;
uint16_t alloc_resp_port, other_port, mail_type, state_reason;
char *nodes = NULL, *partition = NULL, *name = NULL, *resp_host = NULL;
char *account = NULL, *network = NULL, *mail_user = NULL;
@@ -577,6 +580,7 @@ static int _load_job_state(Buf buffer)
safe_unpack16(&kill_on_step_done, buffer);
safe_unpack16(&batch_flag, buffer);
safe_unpack16(&mail_type, buffer);
+ safe_unpack16(&qos, buffer);
safe_unpack16(&state_reason, buffer);
safe_unpackstr_xmalloc(&resp_host, &name_len, buffer);
@@ -662,61 +666,69 @@ static int _load_job_state(Buf buffer)
goto unpack_error;
}
- job_ptr->assoc_id = assoc_id;
- job_ptr->user_id = user_id;
- job_ptr->group_id = group_id;
- job_ptr->time_limit = time_limit;
- job_ptr->priority = priority;
+ xfree(job_ptr->account);
+ job_ptr->account = account;
+ account = NULL; /* reused, nothing left to free */
+ xfree(job_ptr->alloc_node);
+ job_ptr->alloc_node = alloc_node;
+ alloc_node = NULL; /* reused, nothing left to free */
+ job_ptr->alloc_resp_port = alloc_resp_port;
job_ptr->alloc_sid = alloc_sid;
- job_ptr->start_time = start_time;
+ job_ptr->assoc_id = assoc_id;
+ job_ptr->batch_flag = batch_flag;
+ xfree(job_ptr->comment);
+ job_ptr->comment = comment;
+ comment = NULL; /* reused, nothing left to free */
+ job_ptr->db_index = db_index;
job_ptr->end_time = end_time;
- job_ptr->suspend_time = suspend_time;
- job_ptr->pre_sus_time = pre_sus_time;
- job_ptr->tot_sus_time = tot_sus_time;
+ job_ptr->exit_code = exit_code;
+ job_ptr->group_id = group_id;
job_ptr->job_state = job_state;
+ job_ptr->kill_on_node_fail = kill_on_node_fail;
+ job_ptr->kill_on_step_done = kill_on_step_done;
+ xfree(job_ptr->licenses);
+ job_ptr->licenses = licenses;
+ licenses = NULL; /* reused, nothing left to free */
+ job_ptr->mail_type = mail_type;
+ xfree(job_ptr->mail_user);
+ job_ptr->mail_user = mail_user;
+ mail_user = NULL; /* reused, nothing left to free */
+ xfree(job_ptr->name); /* in case duplicate record */
+ job_ptr->name = name;
+ name = NULL; /* reused, nothing left to free */
+ xfree(job_ptr->network);
+ job_ptr->network = network;
+ network = NULL; /* reused, nothing left to free */
job_ptr->next_step_id = next_step_id;
- job_ptr->exit_code = exit_code;
- job_ptr->state_reason = state_reason;
- job_ptr->num_procs = num_procs;
- job_ptr->total_procs = total_procs;
- job_ptr->db_index = db_index;
- job_ptr->assoc_id = assoc_id;
- job_ptr->time_last_active = time(NULL);
- job_ptr->name = name;
- name = NULL; /* reused, nothing left to free */
- xfree(job_ptr->nodes);
- job_ptr->nodes = nodes;
- nodes = NULL; /* reused, nothing left to free */
+ xfree(job_ptr->nodes); /* in case duplicate record */
+ job_ptr->nodes = nodes;
+ nodes = NULL; /* reused, nothing left to free */
if (nodes_completing) {
xfree(job_ptr->nodes_completing);
job_ptr->nodes_completing = nodes_completing;
nodes_completing = NULL; /* reused, nothing left to free */
}
- xfree(job_ptr->alloc_node);
- job_ptr->alloc_node = alloc_node;
- alloc_node = NULL; /* reused, nothing left to free */
- job_ptr->partition = partition;
- partition = NULL; /* reused, nothing left to free */
- job_ptr->account = account;
- account = NULL; /* reused, nothing left to free */
- job_ptr->comment = comment;
- comment = NULL; /* reused, nothing left to free */
- job_ptr->network = network;
- network = NULL; /* reused, nothing left to free */
+ job_ptr->num_procs = num_procs;
+ job_ptr->other_port = other_port;
+ xfree(job_ptr->partition);
+ job_ptr->partition = partition;
+ partition = NULL; /* reused, nothing left to free */
job_ptr->part_ptr = part_ptr;
- job_ptr->kill_on_node_fail = kill_on_node_fail;
- job_ptr->kill_on_step_done = kill_on_step_done;
- job_ptr->batch_flag = batch_flag;
- job_ptr->resp_host = resp_host;
- resp_host = NULL; /* reused, nothing left to free */
- job_ptr->alloc_resp_port = alloc_resp_port;
- job_ptr->other_port = other_port;
- job_ptr->licenses = licenses;
- licenses = NULL; /* reused, nothing left to free */
- job_ptr->mail_type = mail_type;
- job_ptr->mail_user = mail_user;
- mail_user = NULL; /* reused, nothing left to free */
+ job_ptr->pre_sus_time = pre_sus_time;
+ job_ptr->priority = priority;
+ job_ptr->qos = qos;
+ xfree(job_ptr->resp_host);
+ job_ptr->resp_host = resp_host;
+ resp_host = NULL; /* reused, nothing left to free */
job_ptr->select_jobinfo = select_jobinfo;
+ job_ptr->start_time = start_time;
+ job_ptr->state_reason = state_reason;
+ job_ptr->suspend_time = suspend_time;
+ job_ptr->time_last_active = time(NULL);
+ job_ptr->time_limit = time_limit;
+ job_ptr->total_procs = total_procs;
+ job_ptr->tot_sus_time = tot_sus_time;
+ job_ptr->user_id = user_id;
info("recovered job id %u", job_id);
safe_unpack16(&step_flag, buffer);
@@ -2558,7 +2570,27 @@ _copy_job_desc_to_job_record(job_desc_msg_t * job_desc,
job_ptr->account = xstrdup(job_desc->account);
job_ptr->network = xstrdup(job_desc->network);
job_ptr->comment = xstrdup(job_desc->comment);
-
+ if (!wiki_sched_test) {
+ char *sched_type = slurm_get_sched_type();
+ if (strcmp(sched_type, "sched/wiki") == 0)
+ wiki_sched = true;
+ xfree(sched_type);
+ wiki_sched_test = true;
+ }
+ if (wiki_sched && strstr(job_ptr->comment, "QOS:")) {
+ if (strstr(job_ptr->comment, "QOS:execmptall"))
+ job_ptr->qos = QOS_EXEMPT_ALL;
+ else if (strstr(job_ptr->comment, "QOS:execmptwclimit"))
+ job_ptr->qos = QOS_EXEMPT_WC_LIMIT;
+ else if (strstr(job_ptr->comment, "QOS:expedite"))
+ job_ptr->qos = QOS_EXPEDITE;
+ else if (strstr(job_ptr->comment, "QOS:normal"))
+ job_ptr->qos = QOS_NORMAL;
+ else if (strstr(job_ptr->comment, "QOS:standby"))
+ job_ptr->qos = QOS_STANDBY;
+ else
+ job_ptr->qos = QOS_DEFAULT;
+ }
if (job_desc->priority != NO_VAL) /* already confirmed submit_uid==0 */
job_ptr->priority = job_desc->priority;
else {
@@ -2747,9 +2779,6 @@ static void _job_timed_out(struct job_record *job_ptr)
static int _validate_job_desc(job_desc_msg_t * job_desc_msg, int allocate,
uid_t submit_uid)
{
- static bool wiki_sched = false;
- static bool wiki_sched_test = false;
-
/* Permit normal user to specify job id only for sched/wiki
* (Maui scheduler). This was also required with earlier
* versions of the Moab scheduler (wiki2), but was fixed
@@ -3829,6 +3858,21 @@ int update_job(job_desc_msg_t * job_specs, uid_t uid)
job_specs->comment = NULL; /* Nothing left to free */
info("update_job: setting comment to %s for job_id %u",
job_ptr->comment, job_specs->job_id);
+
+ if (wiki_sched && strstr(job_ptr->comment, "QOS:")) {
+ if (strstr(job_ptr->comment, "QOS:execmptall"))
+ job_ptr->qos = QOS_EXEMPT_ALL;
+ else if (strstr(job_ptr->comment, "QOS:execmptwclimit"))
+ job_ptr->qos = QOS_EXEMPT_WC_LIMIT;
+ else if (strstr(job_ptr->comment, "QOS:expedite"))
+ job_ptr->qos = QOS_EXPEDITE;
+ else if (strstr(job_ptr->comment, "QOS:normal"))
+ job_ptr->qos = QOS_NORMAL;
+ else if (strstr(job_ptr->comment, "QOS:standby"))
+ job_ptr->qos = QOS_STANDBY;
+ else
+ job_ptr->qos = QOS_DEFAULT;
+ }
}
if (job_specs->name) {
diff --git a/src/slurmctld/slurmctld.h b/src/slurmctld/slurmctld.h
index 100ff80f14f4b74f7494514d1f11259b2f240166..99db1b91b27a5a7a852ad91bffd997cd33addc7c 100644
--- a/src/slurmctld/slurmctld.h
+++ b/src/slurmctld/slurmctld.h
@@ -274,6 +274,18 @@ extern struct part_record *default_part_loc; /* default partition ptr */
\*****************************************************************************/
extern time_t last_job_update; /* time of last update to part records */
+/* Used for Moab
+ * These QOS values only apply to LLNL's configuration
+ * Other values may apply at other sites */
+typedef enum {
+ QOS_DEFAULT = 0, /* if none set */
+ QOS_EXEMPT_ALL,
+ QOS_EXEMPT_WC_LIMIT,
+ QOS_EXPEDITE,
+ QOS_NORMAL,
+ QOS_STANDBY
+} job_qos_t;
+
#define DETAILS_MAGIC 0xdea84e7
#define JOB_MAGIC 0xf0b7392c
#define STEP_MAGIC 0xce593bc1
@@ -395,6 +407,7 @@ struct job_record {
time_t pre_sus_time; /* time job ran prior to last suspend */
uint32_t priority; /* relative priority of the job,
* zero == held (don't initiate) */
+ uint16_t qos; /* quality of service, used only by Moab */
uint32_t requid; /* requester user ID */
char *resp_host; /* host for srun communications */
select_jobinfo_t select_jobinfo;/* opaque data */