Masternode: Fixing bug on the disconnection and reconnection flow, happens when the controller needs to broadcast a new start message, the remote masternode was rejecting it because thinks that already has it enabled (when the MN is in un-capable state and it was kicked from the network list), loosing the reactivation opportunity and ending in a forever non-capable state.

furszy · furszy · commit 498315700311 · 2020-11-05T10:55:13.000-03:00
diff --git a/src/masternode.cpp b/src/masternode.cpp
@@ -558,11 +558,16 @@ bool CMasternodeBroadcast::CheckInputsAndAdd(int& nDoS)
     // we are a masternode with the same vin (i.e. already activated) and this mnb is ours (matches our Masternode privkey)
     // so nothing to do here for us
     if (fMasterNode && activeMasternode.vin != nullopt &&
-            vin.prevout == activeMasternode.vin->prevout && pubKeyMasternode == activeMasternode.pubKeyMasternode)
+            vin.prevout == activeMasternode.vin->prevout &&
+            pubKeyMasternode == activeMasternode.pubKeyMasternode &&
+            activeMasternode.GetStatus() == ACTIVE_MASTERNODE_STARTED) {
         return true;
+    }
 
     // incorrect ping or its sigTime
-    if(lastPing.IsNull() || !lastPing.CheckAndUpdate(nDoS, false, true)) return false;
+    if(lastPing.IsNull() || !lastPing.CheckAndUpdate(nDoS, false, true)) {
+        return false;
+    }
 
     // search existing Masternode list
     CMasternode* pmn = mnodeman.Find(vin);
