Don't reset storage when rescheduling for collisions

When rescheduling to avoid a collision with an already inflight push, do
not reset the persistent storage as this was not started to begin with
as the "runway" was denied. This avoids extra bogus FileNotFound log
messages.

Bug: Issue 12073
Change-Id: I99cc9e2346ea942b91672ff478605ea9e79d4ad3
diff --git a/src/main/java/com/googlesource/gerrit/plugins/replication/Destination.java b/src/main/java/com/googlesource/gerrit/plugins/replication/Destination.java
index 467db2c..53fb392 100644
--- a/src/main/java/com/googlesource/gerrit/plugins/replication/Destination.java
+++ b/src/main/java/com/googlesource/gerrit/plugins/replication/Destination.java
@@ -537,7 +537,6 @@
         pending.put(uri, pushOp);
         switch (reason) {
           case COLLISION:
-            replicationTasksStorage.get().reset(pushOp);
             @SuppressWarnings("unused")
             ScheduledFuture<?> ignored =
                 pool.schedule(pushOp, config.getRescheduleDelay(), TimeUnit.SECONDS);