Hyukjin Kwon created SPARK-61150:
------------------------------------
Summary: Plan IDs are per session, so a column of another
session's DataFrame resolves to a node of this one
Key: SPARK-61150
URL: https://issues.apache.org/jira/browse/SPARK-61150
Project: Spark
Issue Type: Bug
Components: Connect
Affects Versions: connect-rust-4.3.0
Reporter: Hyukjin Kwon
Each SparkSession numbered its plans from 1, so a Column of a DataFrame of
another session (spark.newSession()) carried a plan ID that also named a node
of this session's plan, and resolved silently to the same-named column.
a.select(b.id) returned [Row(id=1)] and a.filter(b.id == 2).count() returned 0,
where PySpark raises AnalysisException CANNOT_RESOLVE_DATAFRAME_COLUMN. PySpark
numbers plans with one counter of the process (LogicalPlan._nextPlanId).
--
This message was sent by Atlassian Jira
(v10.3.26#10030026)
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
