Details
-
Wish
-
Status: In Progress
-
Minor
-
Resolution: Unresolved
-
2.4.5
-
None
-
None
Description
Our service dynamically creates short-lived YARN clusters in cloud. Spark applications run on these dynamically created clusters. Log data for these applications is stored on a remote file-system. We want a static instance of SparkHistoryServer to view information on jobs that ran on these clusters. We use glob because we cannot have a static list of directories where the log files reside.