[ONOS-7754] Re-populate failure detector with existing node states to avoid failure detections following component configuration change in ClusterStore

Change-Id: Iae27489f8a59da5461370e520646e8734c0303d3
diff --git a/core/store/dist/src/main/java/org/onosproject/store/cluster/impl/DistributedClusterStore.java b/core/store/dist/src/main/java/org/onosproject/store/cluster/impl/DistributedClusterStore.java
index 306ebcb..408c13a 100644
--- a/core/store/dist/src/main/java/org/onosproject/store/cluster/impl/DistributedClusterStore.java
+++ b/core/store/dist/src/main/java/org/onosproject/store/cluster/impl/DistributedClusterStore.java
@@ -459,6 +459,11 @@
         this.minStandardDeviationMillis = minStandardDeviationMillis;
         try {
             failureDetector = new PhiAccrualFailureDetector(minStandardDeviationMillis);
+            nodeStates.forEach((nodeId, state) -> {
+                if (state != State.INACTIVE) {
+                    failureDetector.report(nodeId);
+                }
+            });
         } catch (IllegalArgumentException e) {
             log.warn(e.getMessage());
             this.minStandardDeviationMillis = DEFAULT_MIN_STANDARD_DEVIATION_MILLIS;