Update monitor-all.sh

Sorry before my last request.
I had completely overlooked the syntax.
I have now corrected the syntax and checked the script on a test instance. I hope I have not overlooked anything.

I have extended the script and fixed some bugs. The most important one:

    The detection of whether "onboot" should be started was defective.
    The LCX or VM containers are restarted a maximum of 3 times.
    The detection of the IP address had a bug, the variable was not reset, so if none was found, the last determined one was used.
    I have improved the log output.
    Additionally I added an email output in case of restarts. Email is disabled by default.

misc/monitor-all.sh

@@ -26,67 +26,116 @@ while true; do
 done
 
 echo '#!/usr/bin/env bash
+# Maximum number of restarts
+maxrestartcount=3
+
+# Emails for restarts
+mailaddress='\''admin@domain.tld'\''
+mail="false"
+
+# Search string for IP address, here more precise selection, IP addresses can be excluded.
+searchip='\''192\.168\.|10\.'\''
+
 # Read excluded instances from command line arguments
 excluded_instances=("$@")
-echo "Excluded instances: ${excluded_instances[@]}"
+echo "$(date +'\''%Y-%m-%d %H:%M:%S'\''): Excluded instances: ${excluded_instances[@]}"
 
 while true; do
 
-  for instance in $(pct list | awk '\''{if(NR>1) print $1}'\''; qm list | awk '\''{if(NR>1) print $1}'\''); do
-    # Skip excluded instances
-    if [[ " ${excluded_instances[@]} " =~ " ${instance} " ]]; then
-      echo "Skipping $instance because it is excluded"
-      continue
+  for vmid in $(pct list | awk '\''{if(NR>1) print $1}'\''; qm list | awk '\''{if(NR>1) print $1}'\'')
+  do
+    IP=
+    skip="false"
+    if [ -f /tmp/$vmid.count ] ; then
+      count=$(cat /tmp/$vmid.count)
+    else
+      count=0
     fi
-
     # Determine the type of the instance (container or virtual machine)
-    if pct status $instance >/dev/null 2>&1; then
+    if pct status $vmid >/dev/null 2>&1; then
       # It is a container
       config_cmd="pct config"
-      IP=$(pct exec $instance ip a s dev eth0 | awk '\''/inet / {print $2}'\'' | cut -d/ -f1)
+      test=$(pct status $vmid | grep -q "status: running")
+      if [ $? -eq 0 ] ; then
+        IP=$(pct exec $vmid ip a s dev eth0 | awk '\''/inet / {print $2}'\'' | cut -d/ -f1)
+      fi
     else
       # It is a virtual machine
       config_cmd="qm config"
-      IP=$(qm guest cmd $instance network-get-interfaces | egrep -o "([0-9]{1,3}\.){3}[0-9]{1,3}" | grep -E "192\.|10\." | head -n 1)
+      test=$(qm status $vmid | grep -q "status: running")
+      if [ $? -eq 0 ] ; then
+        IP=$(qm guest cmd $vmid network-get-interfaces | egrep -o "([0-9]{1,3}\.){3}[0-9]{1,3}" | grep -E "$searchip" | head -n 1)
+      fi
     fi
-
+    name=$($config_cmd $vmid | grep name: | awk '\''{ print $2 }'\'')
     # Skip instances based on onboot and templates
-    onboot=$($config_cmd $instance | grep onboot | grep -q "onboot: 0" && echo "true" || echo "false")
-    template=$($config_cmd $instance | grep template | grep -q "template:" && echo "true" || echo "false")
-
-    if [ "$onboot" == "true" ]; then
-      echo "Skipping $instance because it is set not to boot"
-      continue
-    elif [ "$template" == "true" ]; then
-      echo "Skipping $instance because it is a template"
-      continue
+    test=$($config_cmd $vmid | grep "onboot" | awk '\''{ print $2 }'\'')
+    if [ "$test" == "1" ] ; then
+      onboot="true"
+    else
+      onboot="false"
+    fi
+    test=$($config_cmd $vmid | grep "template:" | awk '\''{ print $2 }'\'')
+    if [ "$test" == "1" ] ; then
+      template="true"
+    else
+      template="false"
     fi
 
-    # Ping the instance
-    if ! ping -c 1 $IP >/dev/null 2>&1; then
-      # If the instance can not be pinged, stop and start it
-      if pct status $instance >/dev/null 2>&1; then
-        # It is a container
-        echo "$(date): CT $instance is not responding, restarting..."
-        pct stop $instance >/dev/null 2>&1
-        sleep 5
-        pct start $instance >/dev/null 2>&1
-      else
-        # It is a virtual machine
-        if qm status $instance | grep -q "status: running"; then
-          echo "$(date): VM $instance is not responding, restarting..."
-          qm stop $instance >/dev/null 2>&1
-          sleep 5
+    if [ "$onboot" == "false" ]; then
+      echo "$(date +'\''%Y-%m-%d %H:%M:%S'\''): Skipping $vmid $name because it is set not to boot"
+      skip="true"
+    fi
+    if [ "$template" == "true" ]; then
+      echo "$(date +'\''%Y-%m-%d %H:%M:%S'\''): Skipping $vmid $name because it is a template"
+      skip="true"
+    fi
+    if [ "$skip" == "false" ] ; then
+      # Ping the instance
+      if ! ping -c 1 $IP >/dev/null 2>&1; then
+        if [ $count -le $maxrestartcount ] ; then
+          count=$((count + 1))
+          # If the instance can not be pinged, stop and start it
+          if pct status $vmid >/dev/null 2>&1; then
+            # It is a container
+            echo "$(date +'\''%Y-%m-%d %H:%M:%S'\''): CT $vmid $name is not responding, restarting..."
+            if [ "$mail" == "true" ] ;then
+              echo "CT $vmid $name is not responding, restarting" | mail -s "$(date +'\''%Y-%m-%d %H:%M:%S'\''): $(hostname) - $name" $mailaddress
+            fi
+            pct stop $vmid >/dev/null 2>&1
+            sleep 5
+            pct start $vmid >/dev/null 2>&1
+          else
+            # It is a virtual machine
+            test=$(qm status $vmid | grep -q "status: running")
+            if [ $? -eq 0 ] ; then
+              echo "$(date +'\''%Y-%m-%d %H:%M:%S'\''): VM $vmid $name is not responding, restarting..."
+              if [ "$mail" == "true" ] ;then
+                echo "VM $vmid $name is not responding, restarting" | mail -s "$(date +'\''%Y-%m-%d %H:%M:%S'\''): $(hostname) - $name" $mailaddress
+              fi
+              qm stop $vmid >/dev/null 2>&1
+              sleep 5
+            else
+              echo "$(date +'%Y-%m-%d %H:%M:%S'): VM $vmid $name is not running, starting..."
+            fi
+            qm start $vmid >/dev/null 2>&1
+            echo "$count" > /tmp/$vmid.count
+          fi
         else
-          echo "$(date): VM $instance is not running, starting..."
+          echo "$(date +'\''%Y-%m-%d %H:%M:%S'\''): VM $vmid $name max restart count $count reached"
+          if [ "$mail" == "true" ] ;then
+            echo "VM $vmid $name max restart count $count reached" | mail -s "$(date +'\''%Y-%m-%d %H:%M:%S'\''): $(hostname) - $name" $mailaddress
+          fi
         fi
-        qm start $instance >/dev/null 2>&1
+      else
+        echo "$(date +'\''%Y-%m-%d %H:%M:%S'\''): CT/VM $vmid $name with ip $IP is pingable..."
+        echo "0" > /tmp/$vmid.count
       fi
     fi
   done
 
   # Wait for 5 minutes. (Edit to your needs)
-  echo "$(date): Pausing for 5 minutes..."
+  echo "$(date +'\''%Y-%m-%d %H:%M:%S'\''): Pausing for 5 minutes..."
   sleep 300
 done >/var/log/ping-instances.log 2>&1' >/usr/local/bin/ping-instances.sh
 touch /var/log/ping-instances.log