Merge branch 'latest' into develop

source/installation/add-cluster-config.rst

@@ -82,6 +82,7 @@ client binaries.
    resource-manager/systemd
    resource-manager/bin-override-example
    resource-manager/test
+   resource-manager/advanced-configs
 
 .. _puppet module: https://forge.puppet.com/modules/osc/openondemand
 .. _ansible role: https://galaxy.ansible.com/osc/open_ondemand

source/installation/resource-manager/advanced-configs.rst

@@ -0,0 +1,42 @@
+.. _advanced-resource-manager-configs:
+
+Advanced Resource Manager Configrations
+=======================================
+
+This page details advanced settings for any resource manager
+that could be helpful in using Open OnDemand.
+
+Visualization Nodes
+-------------------
+
+At OSC we offer ``visualization`` nodes that attach the X11 session
+to a GPU so that any rendering can occur on the GPU itself.
+
+We've defined ``vis`` as a `Slurm GRES`_ that jobs can request (i.e.,
+some batch connect applications request this `Slurm GRES`_).
+
+We use a `Slurm prolog`_ to do the actual work of attaching the X11
+session to the GPU.
+
+Other resource managers may have a similar facility for doing work 
+just before the job starts running.
+
+This is an example of that `Slurm prolog`_. 
+
+.. code-block:: bash
+
+  if [[ "$SLURM_LOCALID" == "0" && "$SLURM_JOB_GRES" == *"vis"* ]]; then
+    if [ -n "$CUDA_VISIBLE_DEVICES" ]; then
+      FIRSTGPU=$(echo $CUDA_VISIBLE_DEVICES | tr ',' "\n" | head -1)
+      setsid /usr/bin/X :${FIRSTGPU} -noreset >& /dev/null &
+      sleep 2
+      if [ -n "$DISPLAY" ]; then
+        echo "export OLDDISPLAY=$DISPLAY"
+      fi
+      echo "export DISPLAY=:$FIRSTGPU"
+    fi
+  fi
+
+
+.. _Slurm prolog: https://slurm.schedmd.com/prolog_epilog.html
+.. _Slurm GRES: https://slurm.schedmd.com/gres.html