[
https://issues.apache.org/jira/browse/SPARK-40802?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=17619124#comment-17619124
]
Mingli Rui commented on SPARK-40802:
------------------------------------
Hi, [~hyukjin.kwon], Could you please explain more about +We could probably
introduce a dialect to optimize this further.+ Do you mean let's move
{{JDBCRDD.getQueryOutputSchema}} to a function for JdbcDialect? So that every
concrete Jdbc dialect class has a chance to resolve the schema by their own
way? For example,
{code:java}
abstract class JdbcDialect extends Serializable with Logging {
def getQueryOutputSchema(query: String, options: JDBCOptions): StructType
}
private object MsSqlServerDialect extends JdbcDialect {
override def getQueryOutputSchema(query: String, options: JDBCOptions):
StructType = {
// The provider specific solution
}
}{code}
> Enhance JDBC Connector to use PreparedStatement.getMetaData() to resolve
> schema instead of PreparedStatement.executeQuery()
> ---------------------------------------------------------------------------------------------------------------------------
>
> Key: SPARK-40802
> URL: https://issues.apache.org/jira/browse/SPARK-40802
> Project: Spark
> Issue Type: Improvement
> Components: SQL
> Affects Versions: 2.1.0
> Reporter: Mingli Rui
> Priority: Major
> Original Estimate: 24h
> Remaining Estimate: 24h
>
> Currently, Spark JDBC Connector uses *PreparedStatement.executeQuery()* to
> resolve the JDBCRelation's schema. The schema query is like *s"SELECT * FROM
> $table_or_query WHERE 1=0".*
> But it is not necessary to execute the query. It's enough to *prepare* the
> query. With preparing the statement, the query is parsed and compiled, but is
> not executed. It will be more efficient.
> So, it's better to use PreparedStatement.getMetaData() to resolve schema.
--
This message was sent by Atlassian Jira
(v8.20.10#820010)
---------------------------------------------------------------------
To unsubscribe, e-mail: issues-unsubscr...@spark.apache.org
For additional commands, e-mail: issues-h...@spark.apache.org
