[
https://issues.apache.org/jira/browse/SPARK-31154?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]
Apache Spark reassigned SPARK-31154:
------------------------------------
Assignee: Apache Spark
> Expose basic write metrics for InsertIntoDataSourceCommand
> ----------------------------------------------------------
>
> Key: SPARK-31154
> URL: https://issues.apache.org/jira/browse/SPARK-31154
> Project: Spark
> Issue Type: Improvement
> Components: SQL
> Affects Versions: 3.1.0
> Reporter: Lantao Jin
> Assignee: Apache Spark
> Priority: Major
>
> Spark provides interface `InsertableRelation` and the
> `InsertIntoDataSourceCommand` to delegate the inserting processing to a data
> source. Unlike `DataWritingCommand`, the metrics in
> InsertIntoDataSourceCommand is empty and has no chance to update. So we
> cannot get "number of written files" or "number of output rows" from its
> metrics.
> For example, if a table is a Spark parquet table. We can get the writing
> metrics by:
> {code}
> val df = sql("INSERT INTO TABLE test_table SELECT 1, 'a'")
> val numFiles = df.queryExecution.sparkPlan.metrics("numFiles").value
> {code}
> But if it is a Delta table, we cannot.
--
This message was sent by Atlassian Jira
(v8.3.4#803005)
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
