Hi, Flink版本:1.12 部署模式:on yarn per-job 开发方式:DataStream Api 状态后端:RocksDB Job逻辑为一个15分钟的窗口计算,任务在运行一段时间后会出现内存使用超限,container被yarn kill的现象,目前有不少任务都会存在类似问题: Closing TaskExecutor connection container_e02_1654567136606_1034_01_000012 because: [2023-03-02 08:12:44.794]Container [pid=11455,containerID=container_e02_1654567136606_1034_01_000012] is running 745472B beyond the 'PHYSICAL' memory limit. Current usage: 8.0 GB of 8 GB physical memory used; 10.0 GB of 40 GB virtual memory used. Killing container. 请问: 该如何排查及优化
| | 吴先生 | | 15951914...@163.com |