Fixed missing environment variables in power up/down scripts

stackhpc · Aug 17, 2023 · a731c60 · a731c60
1 parent 63795d3
commit a731c60
Show file tree

Hide file tree

Showing 2 changed files with 14 additions and 7 deletions.
diff --git a/image/k8s-slurmd-create b/image/k8s-slurmd-create
@@ -2,13 +2,15 @@
 
 echo "$(date) Resume invoked $0 $*" &>> /var/log/slurm/power_save.log
 
-echo "Arguments: $* $0 $1" &>> /var/log/slurm/power_save.log
+APISERVER=https://kubernetes.default.svc
+SERVICEACCOUNT=/var/run/secrets/kubernetes.io/serviceaccount
+NAMESPACE=$(cat ${SERVICEACCOUNT}/namespace)
+TOKEN=$(cat ${SERVICEACCOUNT}/token)
+CACERT=${SERVICEACCOUNT}/ca.crt
 
 hosts=$(scontrol show hostnames $1) # this is purely a textual expansion, doens't depend on defined nodes
-echo "Powering up hosts: $hosts" &>> /var/log/slurm/power_save.log
 for host in $hosts
 do
-   echo "Creating $host" &>> /var/log/slurm/power_save.log
-   ( sed s/SLURMD_NODENAME/$host/ /etc/slurm/slurmd-pod-template.yml | kubectl create -f - ) &>> /var/log/slurm/power_save.log
-   echo "done" &>> /var/log/slurm/power_save.log
+   ( sed s/SLURMD_NODENAME/$host/ /etc/slurm/slurmd-pod-template.yml | \
+   kubectl --server $APISERVER --token $TOKEN --certificate-authority $CACERT create -f - )
 done
diff --git a/image/k8s-slurmd-delete b/image/k8s-slurmd-delete
@@ -1,10 +1,15 @@
 #!/usr/bin/bash
-set -euo pipefail
 
 echo "$(date) Suspend invoked $0 $*" >> /var/log/slurm/power_save.log
 
+APISERVER=https://kubernetes.default.svc
+SERVICEACCOUNT=/var/run/secrets/kubernetes.io/serviceaccount
+NAMESPACE=$(cat ${SERVICEACCOUNT}/namespace)
+TOKEN=$(cat ${SERVICEACCOUNT}/token)
+CACERT=${SERVICEACCOUNT}/ca.crt
+
 hosts=$(scontrol show hostnames $1) # this is purely a textual expansion, doens't depend on defined nodes
 for host in $hosts
 do
-   kubectl delete pod $host
+   kubectl --server $APISERVER --token $TOKEN --certificate-authority $CACERT delete pod $host
 done