[
https://issues.apache.org/jira/browse/KAFKA-19404?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=17970231#comment-17970231
]
Greg Harris commented on KAFKA-19404:
-------------------------------------
No, both `/var/plugins` and `/var/plugins/nested1,/var/plugins/nested2` will
isolate plugin1 from plugin2+plugin3. At the same time, they will both not
isolate plugin2 from plugin3.
But `/var/plugins/nested1,/var/plugins/nested2` still isn't recommended because
`dependency2and3` would be isolated from plugin2+plugin3, and that would
probably break things.
To rephrase it a different way, `plugin.path` is a list of "level 0"
directories. "level 1" directories are isolated from one another. "level 2" and
deeper directories are completely irrelevant for JAR packaged plugins, but
matter for class hierarchies.
> Connect's plugin.path documentation does not match behavior
> -----------------------------------------------------------
>
> Key: KAFKA-19404
> URL: https://issues.apache.org/jira/browse/KAFKA-19404
> Project: Kafka
> Issue Type: Task
> Components: connect
> Reporter: Mickael Maison
> Priority: Major
>
> The doc currently states:
> {quote}The list should consist of top level directories that include any
> combination of:
> a) directories immediately containing jars with plugins and their dependencies
> b) uber-jars with plugins and their dependencies
> c) directories immediately containing the package directory structure of
> classes of plugins and their dependencies
> {quote}
> In practice plugins are found even if they are in nested directories. For
> example we can set {{plugin.path}} to {{/var/plugins}} and it will find all
> these plugins:
> {noformat}
> /var/plugins
> └── nested1
> └── nested2
> ├── plugin1
> │ └── plugin1.jar
> ├── plugin2
> │ └── plugin2.jar
> ├── plugin3
> └── plugin3.jar
> {noformat}
> I think we should adjust the documentation to match the current behavior
--
This message was sent by Atlassian Jira
(v8.20.10#820010)
