flink 检查点失败，正常运行

查看flink中的task日志：

2024-01-26 16:10:21,887 WARN  org.apache.flink.runtime.state.filesystem.FsCheckpointStreamFactory [] - Could not close the state stream for hdfs://hk-hdfs/flink/checkpoints/table/XXXXX/shared/6a0e4bdf-27dc-48fe-9e30-546d00c5bfe2.

java.io.IOException: Unable to close file because the last block does not have enough number of replicas.

加大namenode中的参数 30 s ->60s 问题解决

  <property>
      <name>dfs.fsck.http.timeout.ms</name>
      <value>60000</value>
      <text>true</text>
    </property>

2、查看hdfs磁盘io，发现磁盘io到达瓶颈，只够任务运行，做检查点的时候，io不够用了。