[
https://issues.apache.org/jira/browse/FLINK-1875?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14504600#comment-14504600
]
ASF GitHub Bot commented on FLINK-1875:
---------------------------------------
Github user fhueske commented on a diff in the pull request:
https://github.com/apache/flink/pull/604#discussion_r28758925
--- Diff: docs/config.md ---
@@ -370,8 +370,7 @@ system, such as */tmp* in Linux systems.
### Configuring TaskManager processing slots
-A processing slot allows Flink to execute a distributed DataSet
transformation, such as a
-data source or a map-transformation.
+A processing slot allows Flink to execute an instance of a job.
--- End diff --
I think the term *instance of a job* might be misleading. The document uses
*operator instance* or *(user) function instance* but IMO it is not clear what
an *instance of a job* is.
I agree that the previous description wasn't good either, but we might try
to come up with something better.
What about "Flink executes a program in parallel by splitting it into
subtasks and scheduling these subtasks to processing slots."?
> Add figure to documentation describing slots and parallelism
> ------------------------------------------------------------
>
> Key: FLINK-1875
> URL: https://issues.apache.org/jira/browse/FLINK-1875
> Project: Flink
> Issue Type: Improvement
> Components: Documentation
> Affects Versions: 0.9
> Reporter: Robert Metzger
> Assignee: Robert Metzger
>
> Our users are still confused how parallelism and slots are connected to each
> other.
> We tried addressing this issue already with FLINK-1679, but I think we also
> need to have a nice picture in our documentation.
> This is too complicated:
> http://ci.apache.org/projects/flink/flink-docs-master/internal_job_scheduling.html
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
