refactor: reduce waiting time for failover

Now all actions will be retried for ever until timeout. So it is ok to use a smaller timeout. If the operation is not complete with in that time, it will be retried anyway.

go-chaos/cmd/cluster.go

@@ -272,7 +272,7 @@ func forceFailover(flags *Flags) error {
 	changeResponse, err := sendScaleRequest(port, brokersInRegion, true, -1)
 	ensureNoError(err)
 
-	timeout := time.Minute * 25
+	timeout := time.Minute * 5
 	err = waitForChange(port, changeResponse.ChangeId, timeout)
 	ensureNoError(err)