Fix potential redundant router FAULT state by dnsmasq

This message may show during redundant router start up:

FAULT (Restarting DNS forwarder and DHCP server: dnsmasq failed!)

This caused by edithost.sh is racy with keepalived process. They both want to
restart dnsmasq.

Even in normal condition, it's very hard to reproduce this bug. Add file lock
for edithost.sh should solve it.

patches/systemvm/debian/config/root/edithosts.sh

@@ -67,6 +67,33 @@ sed -i  /"$2 "/d /etc/hosts
 sed -i  /"$3"/d /etc/hosts
 echo "$2 $3" >> /etc/hosts
 
+locked=0
+if [ $no_redundant -eq 0 ]
+then
+#for redundant router, grap the lock to prevent racy with keepalived process
+LOCK=/tmp/rrouter.lock
+
+# Wait the lock
+for i in `seq 1 5`
+do
+    if [ ! -e $LOCK ]
+    then
+        touch $LOCK
+        locked=1
+        break
+    fi
+    sleep 1
+    logger -t cloud "edithosts: sleep 1 second wait for the redundant router lock"
+done
+
+if [ $locked -eq 0 ]
+then
+    logger -t cloud "edithosts: fail to get the redundant router lock"
+    logger -t cloud "edithosts: keepalived should able to handle the dnsmasq restart"
+    exit
+fi
+fi
+
 # make dnsmasq re-read files
 pid=$(pidof dnsmasq)
 if [ "$pid" != "" ]
@@ -81,4 +108,10 @@ else
   fi
 fi
 
-exit $?
+ret=$?
+if [ $locked -eq 1 ]
+then
+	rm $LOCK
+fi
+
+exit $ret