[
https://issues.apache.org/jira/browse/SPARK-60507?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]
Uroš Bojanić resolved SPARK-60507.
----------------------------------
Fix Version/s: 4.4.0
Resolution: Fixed
Issue resolved by pull request 59748
[https://github.com/apache/spark/pull/59748]
> Reuse schema resolution connection to read JDBC database metadata
> -----------------------------------------------------------------
>
> Key: SPARK-60507
> URL: https://issues.apache.org/jira/browse/SPARK-60507
> Project: Spark
> Issue Type: Improvement
> Components: SQL
> Affects Versions: 4.3.0
> Reporter: Craiu Constantin-Tiberiu
> Assignee: Craiu Constantin-Tiberiu
> Priority: Trivial
> Labels: pull-request-available
> Fix For: 4.4.0
>
>
> {{JDBCRDD.scanTable}} reads metadata through
> {{{}JDBCDatabaseMetadata.fromJDBCConnectionFactory{}}}, which opens a new
> driver-side connection on every call.
> For a V1 JDBC read ({{{}spark.read.jdbc{}}}, {{{}format("jdbc").load(){}}}),
> schema resolution already opens one connection so we can just reuse that one
> isntead of opening a new connection to the DB.
> This is wasteful for sources where connecting is expensive (TLS, auth
> handshakes, connection limits), and it is unnecessary because the metadata
> can be read on the connection used to resolve the schema.
--
This message was sent by Atlassian Jira
(v10.3.26#10030026)
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
