【Flink】Flink 自动重启过程

最新推荐文章于 2024-06-19 00:03:39 发布

九师兄

最新推荐文章于 2024-06-19 00:03:39 发布

阅读量1.1k

点赞数 3

分类专栏：大数据-flink 文章标签： flink java jvm

本文为博主九师兄（QQ:541711153 欢迎来探讨技术）原创文章，未经允许博主不允许转载。

本文链接：https://blog.csdn.net/qq_21383435/article/details/125774092

版权

大数据-flink 专栏收录该内容

837 篇文章 834 订阅 ¥99.90 ¥99.00

订阅专栏

文章目录

1.概述
2. updateTaskExecutionState
3.处理异常
- 3.1 获取要重启边的id
- 3.2 处理异常handleFailure
4.可能要重启task-maybeRestartTasks

1.概述

写这篇文章是因为我今天的一个任务原本好好的，但是本次自动重启了，然后我就注意到有如下日志。


2022-07-13 18:08:10,032 INFO  org.apache.flink.runtime.executiongraph.failover.flip1.RestartPipelinedRegionFailoverStrategy [] - Calculating tasks to restart to recover the failed task b043b53a8f647a99a305de17a14a0a70_1.
2022-07-13 18:08:10,033 INFO  org.apache.flink.runtime.executiongraph.failover.flip1.RestartPipelinedRegionFailoverStrategy [] - 60 tasks should be restarted to recover the failed task b043b53a8f647a99a305de17a14a0a70_1.
2022-07-13 18:08:10,033 INFO  org.apache.flink.runtime.executiongraph.ExecutionGraph       [] - Job ailpha-rule (ad1f62c4003e70d0e49ef843ae1b796d) switched from state RUNNING to RESTARTING.
2022-07-13 18:08:10,033 INFO  org.apache.flink.runtime.executiongraph.ExecutionGraph       [] - Source: common_metric_logSource (1/6

了解本专栏