[
https://issues.apache.org/jira/browse/FLINK-20614?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=17250391#comment-17250391
]
Kezhu Wang commented on FLINK-20614:
------------------------------------
[~chesnay] Sorry, I did not aware of FLINK-19005 before. Yeh, it is dirty. But
if we don't the dirty work, users may reach FLINK-19005 or
[debugging_classloading|https://ci.apache.org/projects/flink/flink-docs-release-1.12/ops/debugging/debugging_classloading.html#unloading-of-dynamically-loaded-classes-in-user-code]
if they use jdbc in session mode, it is probably not a good experience.
[~f.pompermaier] "jdbc is a mess", I agree with [~chesnay] . Tomcat documents
two issues in their
[docs|https://ci.apache.org/projects/tomcat/tomcat85/docs/jndi-datasource-examples-howto.html#DriverManager,_the_service_provider_mechanism_and_memory_leaks],
another one is new drivers will not be auto-discovered by
{{DriverManager.getConnection}}. Probably, the ultimate solution for clients is
using per-job with dedicated resource management cluster especially in
production environment, this should also avoid other class loader issues
documented in
[debugging_classloading|https://ci.apache.org/projects/flink/flink-docs-release-1.12/ops/debugging/debugging_classloading.html#unloading-of-dynamically-loaded-classes-in-user-code].
> Registered sql drivers not deregistered after task finished in session cluster
> ------------------------------------------------------------------------------
>
> Key: FLINK-20614
> URL: https://issues.apache.org/jira/browse/FLINK-20614
> Project: Flink
> Issue Type: Bug
> Components: Connectors / JDBC, Runtime / Task
> Affects Versions: 1.12.0, 1.13.0
> Reporter: Kezhu Wang
> Priority: Major
>
> {{DriverManager}} keeps registered drivers in its internal data structures
> which prevents they from gc after task finished. I confirm it in standalone
> session cluster by observing that {{ChildFirstClassLoader}} could not be
> reclaimed after several {{GC.run}}, it should exist in all session clusters.
> Tomcat documents
> [this|https://ci.apache.org/projects/tomcat/tomcat85/docs/jndi-datasource-examples-howto.html#DriverManager,_the_service_provider_mechanism_and_memory_leaks]
> and fixes/circumvents this with
> [JdbcLeakPrevention|https://github.com/apache/tomcat/blob/master/java/org/apache/catalina/loader/JdbcLeakPrevention.java#L30].
> Should we solve this in runtime ? Or treat it as connector and clients'
> responsibility to solve it using
> {{RuntimeContext.registerUserCodeClassLoaderReleaseHookIfAbsent}} or similar ?
> Personally, it would be nice to solve in runtime as a catch-all to avoid
> memory-leaking and provide consistent behavior to clients cross per-job and
> session mode.
--
This message was sent by Atlassian Jira
(v8.3.4#803005)
