Add functionality to hpe-logcollector.sh

[32trapper]

It would be good to collect the logs locally from where you ran the hpe-logcollector.sh instead of having to go to each node in the cluster so make sure customer is providing all the necessary logs from the start ensuring a faster resolution. Created an example for collecting the node logs:

./local_copy.sh
Collecting newest archive from hpe-csi-node-266tz ...
-> /var/log/hpe-storage-logs-ocp-w-2.lab.ocp.lan-20260112_152015.tar.gz -> saving as ./hpe-storage-logs-ocp-w-2.lab.ocp.lan-20260112_152015.tar.gz
Collecting newest archive from hpe-csi-node-tk7fh ...
-> /var/log/hpe-storage-logs-ocp-w-1.lab.ocp.lan-20260112_152018.tar.gz -> saving as ./hpe-storage-logs-ocp-w-1.lab.ocp.lan-20260112_152018.tar.gz
[root@ocp-svc ~]# ll hpe-storage-logs-ocp-w-*
-rw-r--r-- 1 root root 2.5M Jan 12 11:20 hpe-storage-logs-ocp-w-1.lab.ocp.lan-20260112_152018.tar.gz
-rw-r--r-- 1 root root 3.5M Jan 12 11:20 hpe-storage-logs-ocp-w-2.lab.ocp.lan-20260112_152015.tar.gz

#!/bin/bash
set -euo pipefail

NS=hpe-storage
CONTAINER=hpe-csi-driver

kubectl get pods -n "$NS" -l app=hpe-csi-node \
  -o jsonpath='{range .items[*]}{.metadata.name}{"\n"}{end}' \
| while read -r POD; do
  echo "Collecting newest archive from ${POD} ..."
  # Find the newest matching tarball inside the pod and print its path and basename
  read -r REMOTE_PATH REMOTE_BASE <<EOF
$(kubectl exec -n "$NS" "$POD" -c "$CONTAINER" -- /bin/bash -c '
  ARCHIVE=$(ls -t /var/log/hpe-storage*.tar.gz 2>/dev/null | head -1)
  if [ -z "$ARCHIVE" ]; then
    echo "NOARCHIVE NOARCHIVE"; exit 0
  fi
  echo "$ARCHIVE $(basename "$ARCHIVE")"
')
EOF

  if [ "$REMOTE_PATH" = "NOARCHIVE" ]; then
    echo "  No archive found in ${POD}"
    continue
  fi

  echo "  -> $REMOTE_PATH -> saving as ./$REMOTE_BASE"
  kubectl exec -n "$NS" "$POD" -c "$CONTAINER" -- /bin/bash -c "cat '$REMOTE_PATH'" > "$REMOTE_BASE"
done