For broadcast, Spark will introduce a new runId as SPARK_JOB_GROUP_ID. That cause we can't follow up this statement's whole life by SPARK_JOB_GROUP_ID. This configuration item will be set when this statement was submitted and it will not be changed. You can get the dependency relationship from job or stage to statement by this configuration item in this statement's whole life. Closes #764 from zhang1002/branch-1.2_set-customized-conf. Closes #764 f78f00ed [张宇翔] Format changes ee303e8f [张宇翔] Format changes 72e8a2ca [张宇翔] Format changes 7d135fbf [张宇翔] Add some Static Variable and clean kyuubi.statement.id in the finally block. 4364ef69 [张宇翔] Add Customized Configuration Item: kyuubi.statement.id Authored-by: 张宇翔 <zhang1002@126.com> Signed-off-by: Kent Yao <yao@apache.org> |
||
|---|---|---|
| .. | ||
| kyuubi-download | ||
| kyuubi-spark-monitor | ||
| kyuubi-spark-sql-engine | ||