无法从jupyter集线器连接到spark群集

jchrr9hc  于 2021-05-26  发布在  Spark
关注(0)|答案(0)|浏览(353)

我已经在kubernetes(azure kubernetes)上用bitnami/spark-helm图表创建了spark部署,我可以从主吊舱运行spark作业。我还在同一个集群下创建了jupyter hub部署,我正在尝试连接到集群。尽管我可以使用以下命令连接到群集:

sc = pyspark.SparkContext(appName="test", master="spark://<spark_cluster_URL>:7077")

但是当我执行任何作业或spark函数进行处理时(比如 sc.range(100).count() or reading any file )在spark ui上,它将显示running,但单元格将持续运行无限长的时间。
我从工人那里得到了以下日志:

Executor app-2020121 finished with state EXITED message Command exited with code 1 exitStatus 1
20/12/11 18:21:30 INFO ExternalShuffleBlockResolver: Clean up non-shuffle and non-RDD files associated with the finished executor 2776
20/12/11 18:21:30 INFO ExternalShuffleBlockResolver: Executor is not registered (appId=app-20201211173521-0003, execId=2776)
20/12/11 18:21:30 INFO Worker: Asked to launch executor app-2020121 for test

有人能告诉我出了什么问题吗?

暂无答案!

目前还没有任何答案,快来回答吧!

相关问题