0
我正在測試flink處理不同數據量的性能,所以我需要Job Runtime來記錄和分析。當處理大量數據時,flink日誌丟失
當我使用flink來處理一萬個記錄這樣的小數據集時,我可以得到Job Runtime日誌如下。
07/18/2017 17:41:47 DataSink (collect())(1/1) switched to FINISHED
07/18/2017 17:41:47 Job execution switched to status FINISHED.
Program execution finished
Job with JobID 3f7658725aaae8cd3427d2aad921f2ef has finished.
Job Runtime: 1124 ms
Accumulator Results:
- c28953fb854da74d18dc7c168b988ca2 (java.util.ArrayList) [15433 elements]
但是當我使用弗林克來處理一點點大的數據集像五萬條記錄,我不能讓工作運行信息,如下,和外殼stucked:
07/18/2017 17:49:33 DataSink (collect())(1/1) switched to FINISHED
07/18/2017 17:49:33 Job execution switched to status FINISHED.
有我需要修改的任何配置?
爲什麼當數據集更大時shell會被刪除?
希望有人能回答我doubts.Thanks〜
我增加了akka.framesize,問題解決了。謝謝〜 – lulijun