Added a port of the sge script to check the status of slurm queues

Also added a script to check apstat and report nodes up, down, avail, and in use on a cray XT or XE system.

hpc/cray_nodestat/README

@@ -0,0 +1,17 @@
+Gmetric script pulled from http://ganglia.info/gmetric
+
+If you are the author of the script, and have an updated version, please fork the repo and
+submit a pull request.  For additional information about the repository, please see the
+README in the repository top-level.
+
+Author: Kris Howard and Fabio Verzelloni
+
+Description:
+
+Reports number of nodes up, down, available, and in use as reported by apstat.
+
+Language: Shell
+
+Category: Statistics :: Cluster
+
+Dependencies: awk, apstat

hpc/cray_nodestat/cray_nodestat.sh

@@ -0,0 +1,27 @@
+#!/bin/bash
+
+# Script to collect stats about a Cray XT or XE system.
+# Reports up, down, avail, and in use
+
+/usr/bin/apstat -n | tail -1 | awk '{system("/usr/bin/gmetric -nnode_total -v" $2 " -tuint16 -u"$2"")} \
+{system("/usr/bin/gmetric -nnode_avail -v" $6 " -tuint16 -u"$6"")} \
+{system("/usr/bin/gmetric -nnode_up -v" $3 " -tuint16 -u"$3"")} \
+{system("/usr/bin/gmetric -nnode_down -v" $7 " -tuint16 -u"$7"")} \
+{system("/usr/bin/gmetric -nnode_use -v" $4 " -tuint16 -u"$4"")}'
+
+#########################################################################
+# Previous Iteration
+########################################################################
+
+#NODE_TOTAL=$(/usr/bin/apstat -n | tail -1 | awk '{print $2}')
+#NODE_AVAIL=$(/usr/bin/apstat -n | tail -1 | awk '{print $6}')
+#NODE_UP=$(/usr/bin/apstat -n | tail -1 | awk '{print $3}')
+#NODE_DOWN=$(/usr/bin/apstat -n | tail -1 | awk '{print $7}')
+#NODE_USE=$(/usr/bin/apstat -n | tail -1 | awk '{print $4}')
+
+#$APPS/system/ganglia-3.1.7/bin/gmetric -n  node_total -v $NODE_TOTAL  -t string -u nodes 
+#$APPS/system/ganglia-3.1.7/bin/gmetric -n  node_avail -v $NODE_AVAIL -t float -u nodes 
+#$APPS/system/ganglia-3.1.7/bin/gmetric -n  node_up -v $NODE_UP -t float -u nodes 
+#$APPS/system/ganglia-3.1.7/bin/gmetric -n  node_down -v $NODE_DOWN -t float -u nodes 
+#$APPS/system/ganglia-3.1.7/bin/gmetric -n  node_use -v $NODE_USE -t float -u nodes 
+

hpc/slurm_jobs/README

@@ -0,0 +1,18 @@
+Gmetric script pulled from http://ganglia.info/gmetric
+
+If you are the author of the script, and have an updated version, please fork the repo and
+submit a pull request.  For additional information about the repository, please see the
+README in the repository top-level.
+
+Author: Jesse Becker
+Edited By: Kris Howard
+
+Description:
+
+Reports number of jobs running, queued, and in error states.
+
+Language: Shell
+
+Category: Statistics :: Cluster
+
+Dependencies: awk

hpc/slurm_jobs/slurm_jobs.sh

@@ -0,0 +1,20 @@
+#!/bin/bash
+
+squeue | awk '
+        BEGIN { pending=running=error=0; }
+        ($5 ~ /^PD/)   { pending++; }
+        ($5 ~ /[rRt]/) { running++; }
+        ($5 ~ /E/ )    { error++;   }
+    END {
+                cmd="/usr/bin/gmetric --name slurmq_pending --value "pending" --type uint16";
+                system(cmd);
+                cmd="/usr/bin/gmetric --name slurmq_running --value "running" --type uint16";
+                system(cmd);
+                cmd="/usr/bin/gmetric --name slurmq_error   --value "error" --type uint16";
+                system(cmd);
+                #print "Pending="pending" Running="running" Errors="error;
+        }'
+
+
+exit
+