Chapter 6 example 6-3 watermark ts will overflow when job recovery from failure

[dispensable]

examples-scala/src/main/scala/io/github/streamingwithflink/chapter6/WatermarkGeneration.scala

Lines 45 to 62 in c188681

	class PeriodicAssigner extends AssignerWithPeriodicWatermarks[SensorReading] {
	// 1 min in ms
	val bound: Long = 60 * 1000
	// the maximum observed timestamp
	var maxTs: Long = Long.MinValue
	override def getCurrentWatermark: Watermark = {
	new Watermark(maxTs - bound)
	}
	override def extractTimestamp(r: SensorReading, previousTS: Long): Long = {
	// update maximum timestamp
	maxTs = maxTs.max(r.timestamp)
	// return record timestamp
	r.timestamp
	}
	}

When the task recovery from the cluster failure, the maxTS is initialized with Long.MinValue and the task manager will call getCurrentWatermark method which will - 60 * 1000 and make the maxTS to 9223372036854715808. This invalid ts will block the stream work with the web Dashboard all green but actually they will never push the watermark forward. I know this is just a example but maybe a little warning about this example's job recovery behiver will make the reader less frustrated with their code?