Replies: 8 comments
-
Hello @qshian, |
Beta Was this translation helpful? Give feedback.
-
May be caused by #2686, fixed in 1.5.2 |
Beta Was this translation helpful? Give feedback.
-
@pan3793 我看修复的是在 非connection level的问题,我这个是在 connection level 发生的 |
Beta Was this translation helpful? Give feedback.
-
After a second look, the Spark Application name does not match the Kyuubi App name pattern, are you sure they were created by Kyuubi? |
Beta Was this translation helpful? Give feedback.
-
@pan3793 是的, 在jdbc模式URL后面 加了一个参数 ?spark.app.name =xxxx |
Beta Was this translation helpful? Give feedback.
-
does it occur every time? if yes, it may be caused by your client code, would you please share your JDBC client code? |
Beta Was this translation helpful? Give feedback.
-
the |
Beta Was this translation helpful? Give feedback.
-
@pan3793 经过验证 是 org.apache.commons.dbcp 在创建 datasource的时候 ,有一个效验连接工厂的方法 |
Beta Was this translation helpful? Give feedback.
Uh oh!
There was an error while loading. Please reload this page.
-
Code of Conduct
Search before asking
Describe the bug
用java jdbc的方式连接 kyuubi ,提交一条sql , spark 会生成两个集群。一个是show databases,另外一个是 show databases + 提交的sql 。
这种在connection的模式下, 我的理解,show tables是engine_initializing_queries 在SparkSqlEngine 135 行执行,后面都是返回这个session ,为什么会产生两个spark 作业了,向k8s申请两次不同的资源了。
Affects Version(s)
1.5.1
Kyuubi Server Log Output
Kyuubi Engine Log Output
Kyuubi Server Configurations
Kyuubi Engine Configurations
没有配置
Additional context
无
Are you willing to submit PR?
Beta Was this translation helpful? Give feedback.
All reactions