From 6874ce844a00ce5b6d045b96ba57997427428710 Mon Sep 17 00:00:00 2001
From: Michael Lueken <63728921+MichaelLueken@users.noreply.github.com>
Date: Fri, 9 Feb 2024 15:36:44 -0500
Subject: [PATCH] [develop] Add -n 1 to allow the use of the service partition
 (#1012)

Following the Slurm update on Hera and Jet, the service partition is no longer usable within the SRW App. The necessary changes to allow the service partition to once again function properly have been made, by adding -n 1 to the SCHED_NATIVE_CMD_HPSS variable in the Hera and Jet machine yaml files, and updating the native entry in the parm/wflow/verify_pre.yaml and parm/wflow/aqm_prep.yaml files.
---
 parm/wflow/aqm_prep.yaml   | 1 +
 parm/wflow/verify_pre.yaml | 4 ++++
 ush/machine/hera.yaml      | 1 +
 ush/machine/jet.yaml       | 1 +
 4 files changed, 7 insertions(+)

diff --git a/parm/wflow/aqm_prep.yaml b/parm/wflow/aqm_prep.yaml
index 36d2c8af2a..6cfab161d7 100644
--- a/parm/wflow/aqm_prep.yaml
+++ b/parm/wflow/aqm_prep.yaml
@@ -23,6 +23,7 @@ default_aqm_task: &default_aqm
 task_nexus_gfs_sfc:
   <<: *default_aqm
   command: '&LOAD_MODULES_RUN_TASK_FP; "nexus_gfs_sfc" "&JOBSdir;/JREGIONAL_NEXUS_GFS_SFC"'
+  native: '{% if platform.get("SCHED_NATIVE_CMD_HPSS") %}{{ platform.SCHED_NATIVE_CMD_HPSS }}{% else %}{{ platform.SCHED_NATIVE_CMD}}{% endif %}'
   partition: '{% if platform.get("PARTITION_HPSS") %}&PARTITION_HPSS;{% else %}None{% endif %}'
   join: !cycstr '&LOGDIR;/{{ jobname }}_@Y@m@d@H&LOGEXT;'
   memory: 2G
diff --git a/parm/wflow/verify_pre.yaml b/parm/wflow/verify_pre.yaml
index bc1ca11078..eb1a7eb796 100644
--- a/parm/wflow/verify_pre.yaml
+++ b/parm/wflow/verify_pre.yaml
@@ -30,6 +30,7 @@ task_get_obs_ccpa:
     OBS_DIR: '&CCPA_OBS_DIR;'
     OBTYPE: 'CCPA'
     FHR: '{% for h in range(0, workflow.FCST_LEN_HRS+1) %}{{ " %02d" % h  }}{% endfor %}'
+  native: '{% if platform.get("SCHED_NATIVE_CMD_HPSS") %}{{ platform.SCHED_NATIVE_CMD_HPSS }}{% else %}{{ platform.SCHED_NATIVE_CMD}}{% endif %}'
   partition: '{% if platform.get("PARTITION_HPSS") %}&PARTITION_HPSS;{% else %}None{% endif %}'
   queue: "&QUEUE_HPSS;"
   walltime: 00:45:00
@@ -42,6 +43,7 @@ task_get_obs_nohrsc:
     OBS_DIR: '&NOHRSC_OBS_DIR;'
     OBTYPE: 'NOHRSC'
     FHR: '{% for h in range(0, workflow.FCST_LEN_HRS+1) %}{{ " %02d" % h  }}{% endfor %}'
+  native: '{% if platform.get("SCHED_NATIVE_CMD_HPSS") %}{{ platform.SCHED_NATIVE_CMD_HPSS }}{% else %}{{ platform.SCHED_NATIVE_CMD}}{% endif %}'
   partition: '{% if platform.get("PARTITION_HPSS") %}&PARTITION_HPSS;{% else %}None{% endif %}'
   queue: "&QUEUE_HPSS;"
   walltime: 00:45:00
@@ -55,6 +57,7 @@ task_get_obs_mrms:
     OBTYPE: 'MRMS'
     VAR: 'REFC RETOP'
     FHR: '{% for h in range(0, workflow.FCST_LEN_HRS+1) %}{{ " %02d" % h  }}{% endfor %}'
+  native: '{% if platform.get("SCHED_NATIVE_CMD_HPSS") %}{{ platform.SCHED_NATIVE_CMD_HPSS }}{% else %}{{ platform.SCHED_NATIVE_CMD}}{% endif %}'
   partition: '{% if platform.get("PARTITION_HPSS") %}&PARTITION_HPSS;{% else %}None{% endif %}'
   queue: "&QUEUE_HPSS;"
   walltime: 00:45:00
@@ -68,6 +71,7 @@ task_get_obs_ndas:
     FHR: '{% for h in range(0, workflow.FCST_LEN_HRS+1) %}{{ " %02d" % h  }}{% endfor %}'
   command: '&LOAD_MODULES_RUN_TASK_FP; "get_obs" "&JOBSdir;/JREGIONAL_GET_VERIF_OBS"'
   queue: "&QUEUE_HPSS;"
+  native: '{% if platform.get("SCHED_NATIVE_CMD_HPSS") %}{{ platform.SCHED_NATIVE_CMD_HPSS }}{% else %}{{ platform.SCHED_NATIVE_CMD}}{% endif %}'
   partition: '{% if platform.get("PARTITION_HPSS") %}&PARTITION_HPSS;{% else %}None{% endif %}'
   walltime: 02:00:00
 
diff --git a/ush/machine/hera.yaml b/ush/machine/hera.yaml
index 2ad41e36aa..8d751ae891 100644
--- a/ush/machine/hera.yaml
+++ b/ush/machine/hera.yaml
@@ -21,6 +21,7 @@ platform:
   RUN_CMD_NEXUS: srun -n ${nprocs} --export=ALL
   RUN_CMD_AQMLBC: srun --export=ALL -n ${NUMTS}
   SCHED_NATIVE_CMD: --export=NONE
+  SCHED_NATIVE_CMD_HPSS: -n 1 --export=NONE
   PRE_TASK_CMDS: '{ ulimit -s unlimited; ulimit -a; }'
   TEST_EXTRN_MDL_SOURCE_BASEDIR: /scratch1/NCEPDEV/nems/role.epic/UFS_SRW_data/develop/input_model_data
   TEST_AQM_INPUT_BASEDIR: /scratch1/NCEPDEV/nems/role.epic/UFS_SRW_data/develop/aqm_data
diff --git a/ush/machine/jet.yaml b/ush/machine/jet.yaml
index b5cb19cd0b..93d375ee02 100644
--- a/ush/machine/jet.yaml
+++ b/ush/machine/jet.yaml
@@ -19,6 +19,7 @@ platform:
   RUN_CMD_SERIAL: time
   RUN_CMD_UTILS: srun --export=ALL
   SCHED_NATIVE_CMD: --export=NONE
+  SCHED_NATIVE_CMD_HPSS: -n 1 --export=NONE
   PRE_TASK_CMDS: '{ ulimit -s unlimited; ulimit -a; }'
   TEST_EXTRN_MDL_SOURCE_BASEDIR: /mnt/lfs4/HFIP/hfv3gfs/role.epic/UFS_SRW_data/develop/input_model_data
   TEST_PREGEN_BASEDIR: /mnt/lfs4/HFIP/hfv3gfs/role.epic/UFS_SRW_data/develop/FV3LAM_pregen