How to solve the problem of error reporting in Flink

16:14:43,781 WARN  org.apache.flink.streaming.api.functions.timestamps.AscendingTimestampExtractor  - Timestamp monotony violated: -159096192 < 982926000

I used it

 DataStream<Tuple3<String,String,Integer>> timedData = dataStream
                .assignTimestampsAndWatermarks(new AscendingTimestampExtractor<Tuple3<String,String,Integer>>() {
                    @Override
                    public long extractAscendingTimestamp(Tuple3<String,String,Integer> s) {
                        return s.f2*1000;
                    }
                });

The error content violates the monotonicity of timestamps. The timestamps passed in by the ascendingtimestampextractor method must be monotonous.

After checking the simulation data, it is found that some timestamps are not monotonic, resulting in error.

It’s a time stamp that is smaller than the previous time stamp. Please correct it.

And then I thought, will this problem occur when Kafka is used in the future?Forget it. Let’s talk about it later

Read More: