Occasionally, Flink data is overstocked, resulting in checkpoint failure

One 、 Problem description

Flink Program test succeeded , After running for a while , Data occasionally leads to data backlog checkpoint Failure

Two 、 Question why

Possible causes , Memory and disk have small carrying capacity , Data write delay

3、 ... and 、 Solution （ temporary ）

1. increase checkpoint Creation time , Reduce writes （.enableCheckpointing）
2. Increase the number of failures allowed （.setTolerableCheckpointFailureNumber）

env.setStateBackend(new FsStateBackend(FSSTATEBACKEND));
        env.enableCheckpointing(10000);//  Every time  ** ms  Start once  checkpoint
        env.getCheckpointConfig().setCheckpointingMode(CheckpointingMode.EXACTLY_ONCE);//  Set the mode to precise once 
        env.getCheckpointConfig().setCheckpointTimeout(100000);// Checkpoint  Must be in ** ms Finish in , Otherwise, they will be abandoned 
        env.getCheckpointConfig().setMaxConcurrentCheckpoints(2);//  Only one is allowed at a time  checkpoint  Conduct 
        env.getCheckpointConfig().setMinPauseBetweenCheckpoints(3000);//  confirm  checkpoints  The time between will be  ** ms
        env.getCheckpointConfig().setTolerableCheckpointFailureNumber(5);
        env.setRestartStrategy(RestartStrategies.fixedDelayRestart(3, Time.of(10,TimeUnit.SECONDS)));// Restart strategy ： restart 3 Time , interval 10s