fix(update-lxcs): keep going when one container fails

A failed patchmon-agent report, apk upgrade or pct start ended the whole run under set -e (18 runs in three days). The run now updates the other containers, lists the failed ones at the end and exits 1.

tools/pve/clean-lxcs.sh

@@ -54,35 +54,35 @@ function run_lxc_clean() {
   header_info
   name=$(pct exec "$container" -- hostname)
 
-  pct exec "$container" -- sh -c '
+  pct exec "$container" -- bash -c '
     BL="\033[36m"; GN="\033[1;92m"; CL="\033[m"
     name=$(hostname)
     if [ -e /etc/alpine-release ]; then
-      printf "%b\n\n" "${BL}[Info]${GN} Cleaning $name (Alpine)${CL}"
+      echo -e "${BL}[Info]${GN} Cleaning $name (Alpine)${CL}\n"
       apk cache clean
       find /var/log -type f -delete 2>/dev/null
       find /tmp -mindepth 1 -delete 2>/dev/null
       apk update
     elif [ -e /etc/redhat-release ]; then
-      printf "%b\n\n" "${BL}[Info]${GN} Cleaning $name (CentOS)${CL}"
+      echo -e "${BL}[Info]${GN} Cleaning $name (CentOS)${CL}\n"
       yum clean all
       find /var/log -type f -delete 2>/dev/null
       find /tmp -mindepth 1 -delete 2>/dev/null
       yum update
       yum upgrade -y
     else
-      printf "%b\n\n" "${BL}[Info]${GN} Cleaning $name (Debian/Ubuntu)${CL}"
+      echo -e "${BL}[Info]${GN} Cleaning $name (Debian/Ubuntu)${CL}\n"
       find /var/cache -type f -delete 2>/dev/null
       find /var/log -type f -delete 2>/dev/null
       find /tmp -mindepth 1 -delete 2>/dev/null
       apt -y --purge autoremove
       apt -y autoclean
+      rm -rf /var/lib/apt/lists/*
       apt update
     fi
   '
 }
 
-failed_containers=()
 for container in $(pct list | awk '{if(NR>1) print $1}'); do
   excluded=0
   for ex in $excluded_containers; do
@@ -112,25 +112,17 @@ for container in $(pct list | awk '{if(NR>1) print $1}'); do
 
   if [ "$template" == "false" ] && [ "$status" == "status: stopped" ]; then
     echo -e "${BL}[Info]${GN} Starting${BL} $container ${CL} \n"
-    if ! pct start "$container"; then
-      failed_containers+=("$container (start)")
-      continue
-    fi
+    pct start "$container"
     echo -e "${BL}[Info]${GN} Waiting For${BL} $container${CL}${GN} To Start ${CL} \n"
     sleep 5
-    run_lxc_clean "$container" || failed_containers+=("$container")
+    run_lxc_clean "$container"
     echo -e "${BL}[Info]${GN} Shutting down${BL} $container ${CL} \n"
     pct shutdown "$container" &
   elif [ "$status" == "status: running" ]; then
-    run_lxc_clean "$container" || failed_containers+=("$container")
+    run_lxc_clean "$container"
   fi
 done
 
 wait
 header_info
-if [ "${#failed_containers[@]}" -gt 0 ]; then
-  echo -e "${RD}Cleaning failed in these containers:${CL}"
-  printf '%s\n' "${failed_containers[@]}"
-  exit 1
-fi
 echo -e "${GN} Finished, Selected Containers Cleaned. ${CL} \n"

tools/pve/update-lxcs.sh

@@ -84,6 +84,7 @@ function update_container() {
 }
 
 containers_needing_reboot=()
+failed_containers=()
 header_info
 for container in $(pct list | awk '{if(NR>1) print $1}'); do
   if [[ " ${excluded_containers[@]} " =~ " $container " ]]; then
@@ -101,14 +102,17 @@ for container in $(pct list | awk '{if(NR>1) print $1}'); do
     template=$(pct config $container | grep -q "template:" && echo "true" || echo "false")
     if [ "$template" == "false" ] && [ "$status" == "status: stopped" ]; then
       echo -e "${BL}[Info]${GN} Starting${BL} $container ${CL} \n"
-      pct start $container
+      if ! pct start $container; then
+        failed_containers+=("$container (start)")
+        continue
+      fi
       echo -e "${BL}[Info]${GN} Waiting For${BL} $container${CL}${GN} To Start ${CL} \n"
       sleep 5
-      update_container $container
+      update_container $container || failed_containers+=("$container")
       echo -e "${BL}[Info]${GN} Shutting down${BL} $container ${CL} \n"
       pct shutdown $container &
     elif [ "$status" == "status: running" ]; then
-      update_container $container
+      update_container $container || failed_containers+=("$container")
     fi
     if [ "$status" == "status: running" ]; then
       if pct exec "$container" -- [ -e "/var/run/reboot-required" ]; then
@@ -119,14 +123,20 @@ for container in $(pct list | awk '{if(NR>1) print $1}'); do
       # check if patchmon agent is present in container and run a report if found
       if pct exec "$container" -- [ -e "/usr/local/bin/patchmon-agent" ]; then
         echo -e "${BL}[Info]${GN} patchmon-agent found in ${BL} $container ${CL}, triggering report. \n"
-        pct exec "$container" -- "/usr/local/bin/patchmon-agent" "report"
+        pct exec "$container" -- "/usr/local/bin/patchmon-agent" "report" || echo -e "${YW}[Warn]${CL} patchmon-agent report failed in $container\n"
       fi
     fi
   fi
 done
 wait
 header_info
-echo -e "${GN}The process is complete, and the containers have been successfully updated.${CL}\n"
+if [ "${#failed_containers[@]}" -gt 0 ]; then
+  echo -e "${RD}The update failed in these containers:${CL}"
+  printf '%s\n' "${failed_containers[@]}"
+  echo ""
+else
+  echo -e "${GN}The process is complete, and the containers have been successfully updated.${CL}\n"
+fi
 if [ "${#containers_needing_reboot[@]}" -gt 0 ]; then
   echo -e "${RD}The following containers require a reboot:${CL}"
   for container_name in "${containers_needing_reboot[@]}"; do
@@ -134,3 +144,4 @@ if [ "${#containers_needing_reboot[@]}" -gt 0 ]; then
   done
 fi
 echo ""
+[ "${#failed_containers[@]}" -eq 0 ] || exit 1