From f5d252a8dc0420f4f9f88c352517af2f3d0d94e2 Mon Sep 17 00:00:00 2001
From: Don Lipari <lipari1@llnl.gov>
Date: Tue, 1 Sep 2009 21:38:20 +0000
Subject: [PATCH] Completed work to add qos support to salloc/sbatch/srun
---
doc/html/accounting.shtml | 5 ++++
doc/html/accounting_storageplugins.shtml | 4 +--
doc/html/resource_limits.shtml | 5 ++++
doc/man/man1/salloc.1 | 8 ++++++
doc/man/man1/sbatch.1 | 8 ++++++
doc/man/man1/srun.1 | 8 ++++++
src/api/job_info.c | 35 +++++++++++++++---------
src/slurmctld/job_mgr.c | 6 +++-
8 files changed, 63 insertions(+), 16 deletions(-)
diff --git a/doc/html/accounting.shtml b/doc/html/accounting.shtml
index b1c1031a330..7447c02352e 100644
--- a/doc/html/accounting.shtml
+++ b/doc/html/accounting.shtml
@@ -226,6 +226,11 @@ prevent users from accessing invalid accounts.
<li>limits - This will enforce limits set to associations. By setting
this option, the 'associations' option is also set.
</li>
+<li>qos - This will require all jobs to specify (either overtly or by
+default) a valid qos (Quality of Service). QOS values are defined for
+each association in the database. By setting this option, the
+'associations' option is also set.
+</li>
<li>wckeys - This will prevent users from running jobs under a wckey
that they don't have access to. By using this option, the
'associations' option is also set. The 'TrackWCKey' option is also
diff --git a/doc/html/accounting_storageplugins.shtml b/doc/html/accounting_storageplugins.shtml
index 7bdac1ae532..ebe7221031a 100644
--- a/doc/html/accounting_storageplugins.shtml
+++ b/doc/html/accounting_storageplugins.shtml
@@ -912,8 +912,8 @@ database with.
<dt><span class="commandline">AccountingStorageEnforce</span>
<dd>Specifies if we should enforce certain things be in existence
before allowing job submissions and such valid options are
- "associations, limits, and wckeys". You can use any combination of
- those listed.
+ "associations, limits, qos, and wckeys". You can use any combination
+ of those listed.
</dl>
<h2>Versioning</h2>
diff --git a/doc/html/resource_limits.shtml b/doc/html/resource_limits.shtml
index afcb619afe1..a2e3469ffdf 100644
--- a/doc/html/resource_limits.shtml
+++ b/doc/html/resource_limits.shtml
@@ -52,6 +52,11 @@ prevent users from accessing invalid accounts.
<li>limits - This will enforce limits set to associations. By setting
this option, the 'associations' option is also set.
</li>
+<li>qos - This will require all jobs to specify (either overtly or by
+default) a valid qos (Quality of Service). QOS values are defined for
+each association in the database. By setting this option, the
+'associations' option is also set.
+</li>
<li>wckeys - This will prevent users from running jobs under a wckey
that they don't have access to. By using this option, the
'associations' option is also set. The 'TrackWCKey' option is also
diff --git a/doc/man/man1/salloc.1 b/doc/man/man1/salloc.1
index 27129c3ab9b..32de207275e 100644
--- a/doc/man/man1/salloc.1
+++ b/doc/man/man1/salloc.1
@@ -709,6 +709,14 @@ partition as designated by the system administrator.
\fB\-Q\fR, \fB\-\-quiet\fR
Suppress informational messages from salloc. Errors will still be displayed.
+.TP
+\fB\-\-qos\fR=<\fIqos\fR>
+Request a quality of service for the job. QOS values can be defined
+for each user/cluster/account association in the SLURM database.
+Users will be limited to their association's defined set of qos's when
+the SLURM configuration parameter, AccountingStorageEnforce, is set to
+"qos".
+
.TP
\fB\-\-reservation\fR=<\fIname\fR>
Allocate resources for the job from the named reservation.
diff --git a/doc/man/man1/sbatch.1 b/doc/man/man1/sbatch.1
index 4ba28c46128..e6b7234c4f8 100644
--- a/doc/man/man1/sbatch.1
+++ b/doc/man/man1/sbatch.1
@@ -796,6 +796,14 @@ The maximum stack size
\fB\-Q\fR, \fB\-\-quiet\fR
Suppress informational messages from sbatch. Errors will still be displayed.
+.TP
+\fB\-\-qos\fR=<\fIqos\fR>
+Request a quality of service for the job. QOS values can be defined
+for each user/cluster/account association in the SLURM database.
+Users will be limited to their association's defined set of qos's when
+the SLURM configuration parameter, AccountingStorageEnforce, is set to
+"qos".
+
.TP
\fB\-\-requeue\fR
Specifies that the batch job should be requeued after node failure.
diff --git a/doc/man/man1/srun.1 b/doc/man/man1/srun.1
index cd3266d164c..997f7d1f774 100644
--- a/doc/man/man1/srun.1
+++ b/doc/man/man1/srun.1
@@ -884,6 +884,14 @@ disables the status feature normally available when \fBsrun\fR receives
a single Ctrl\-C and causes \fBsrun\fR to instead immediately terminate the
running job.
+.TP
+\fB\-\-qos\fR=<\fIqos\fR>
+Request a quality of service for the job. QOS values can be defined
+for each user/cluster/account association in the SLURM database.
+Users will be limited to their association's defined set of qos's when
+the SLURM configuration parameter, AccountingStorageEnforce, is set to
+"qos".
+
.TP
\fB\-r\fR, \fB\-\-relative\fR=<\fIn\fR>
Run a job step relative to node \fIn\fR of the current allocation.
diff --git a/src/api/job_info.c b/src/api/job_info.c
index a2679fe7b03..61bf810097b 100644
--- a/src/api/job_info.c
+++ b/src/api/job_info.c
@@ -522,9 +522,8 @@ line7: /****** Line 7 ******/
/****** Line 11 ******/
snprintf(tmp_line, sizeof(tmp_line),
- "Dependency=%s Account=%s Requeue=%u Restarts=%u",
- job_ptr->dependency, job_ptr->account, job_ptr->requeue,
- job_ptr->restart_cnt);
+ "Dependency=%s Requeue=%u Restarts=%u",
+ job_ptr->dependency, job_ptr->requeue, job_ptr->restart_cnt);
xstrcat(out, tmp_line);
if (one_liner)
xstrcat(out, " ");
@@ -532,6 +531,16 @@ line7: /****** Line 7 ******/
xstrcat(out, "\n ");
/****** Line 12 ******/
+ snprintf(tmp_line, sizeof(tmp_line),
+ "Account=%s QOS=%s",
+ job_ptr->account, job_ptr->qos);
+ xstrcat(out, tmp_line);
+ if (one_liner)
+ xstrcat(out, " ");
+ else
+ xstrcat(out, "\n ");
+
+ /****** Line 13 ******/
if (job_ptr->state_desc) {
/* Replace white space with underscore for easier parsing */
for (j=0; job_ptr->state_desc[j]; j++) {
@@ -550,7 +559,7 @@ line7: /****** Line 7 ******/
else
xstrcat(out, "\n ");
- /****** Line 13 ******/
+ /****** Line 14 ******/
snprintf(tmp_line, sizeof(tmp_line), "Req%s=%s Req%sIndices=",
nodelist, job_ptr->req_nodes, nodelist);
xstrcat(out, tmp_line);
@@ -567,7 +576,7 @@ line7: /****** Line 7 ******/
else
xstrcat(out, "\n ");
- /****** Line 14 ******/
+ /****** Line 15 ******/
snprintf(tmp_line, sizeof(tmp_line), "Exc%s=%s Exc%sIndices=",
nodelist, job_ptr->exc_nodes, nodelist);
xstrcat(out, tmp_line);
@@ -584,7 +593,7 @@ line7: /****** Line 7 ******/
else
xstrcat(out, "\n ");
- /****** Line 15 ******/
+ /****** Line 16 ******/
slurm_make_time_str((time_t *)&job_ptr->submit_time, time_str,
sizeof(time_str));
snprintf(tmp_line, sizeof(tmp_line), "SubmitTime=%s ",
@@ -600,7 +609,7 @@ line7: /****** Line 7 ******/
time_str, (long int)job_ptr->pre_sus_time);
xstrcat(out, tmp_line);
- /****** Lines 16, 17 (optional, batch only) ******/
+ /****** Lines 17, 18 (optional, batch only) ******/
if (job_ptr->batch_flag) {
if (one_liner)
xstrcat(out, " ");
@@ -617,7 +626,7 @@ line7: /****** Line 7 ******/
xstrcat(out, tmp_line);
}
- /****** Line 18 (optional) ******/
+ /****** Line 19 (optional) ******/
if (job_ptr->comment) {
if (one_liner)
xstrcat(out, " ");
@@ -628,7 +637,7 @@ line7: /****** Line 7 ******/
xstrcat(out, tmp_line);
}
- /****** Line 19 (optional) ******/
+ /****** Line 20 (optional) ******/
select_g_select_jobinfo_sprint(job_ptr->select_jobinfo,
select_buf, sizeof(select_buf),
SELECT_PRINT_MIXED);
@@ -640,7 +649,7 @@ line7: /****** Line 7 ******/
xstrcat(out, select_buf);
}
#ifdef HAVE_BG
- /****** Line 20 (optional) ******/
+ /****** Line 21 (optional) ******/
select_g_select_jobinfo_sprint(job_ptr->select_jobinfo,
select_buf, sizeof(select_buf),
SELECT_PRINT_BLRTS_IMAGE);
@@ -653,7 +662,7 @@ line7: /****** Line 7 ******/
"BlrtsImage=%s", select_buf);
xstrcat(out, tmp_line);
}
- /****** Line 21 (optional) ******/
+ /****** Line 22 (optional) ******/
select_g_select_jobinfo_sprint(job_ptr->select_jobinfo,
select_buf, sizeof(select_buf),
SELECT_PRINT_LINUX_IMAGE);
@@ -671,7 +680,7 @@ line7: /****** Line 7 ******/
#endif
xstrcat(out, tmp_line);
}
- /****** Line 22 (optional) ******/
+ /****** Line 23 (optional) ******/
select_g_select_jobinfo_sprint(job_ptr->select_jobinfo,
select_buf, sizeof(select_buf),
SELECT_PRINT_MLOADER_IMAGE);
@@ -684,7 +693,7 @@ line7: /****** Line 7 ******/
"MloaderImage=%s", select_buf);
xstrcat(out, tmp_line);
}
- /****** Line 23 (optional) ******/
+ /****** Line 24 (optional) ******/
select_g_select_jobinfo_sprint(job_ptr->select_jobinfo,
select_buf, sizeof(select_buf),
SELECT_PRINT_RAMDISK_IMAGE);
diff --git a/src/slurmctld/job_mgr.c b/src/slurmctld/job_mgr.c
index 6db15a84cb6..534622158ee 100644
--- a/src/slurmctld/job_mgr.c
+++ b/src/slurmctld/job_mgr.c
@@ -4198,7 +4198,11 @@ void pack_job(struct job_record *dump_job_ptr, uint16_t show_flags, Buf buffer)
packstr(dump_job_ptr->comment, buffer);
slurm_mutex_lock(&assoc_mgr_qos_lock);
- packstr(acct_qos_str(assoc_mgr_qos_list, dump_job_ptr->qos), buffer);
+ if (assoc_mgr_qos_list)
+ packstr(acct_qos_str(assoc_mgr_qos_list, dump_job_ptr->qos),
+ buffer);
+ else
+ packstr(NULL, buffer);
slurm_mutex_unlock(&assoc_mgr_qos_lock);
packstr(dump_job_ptr->licenses, buffer);
--
GitLab