[
https://issues.apache.org/jira/browse/BEAM-2007?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=15973797#comment-15973797
]
ASF GitHub Bot commented on BEAM-2007:
--------------------------------------
GitHub user tgroh opened a pull request:
https://github.com/apache/beam/pull/2587
[BEAM-2007] Ensure all Read outputs are consumed in Dataflow
Be sure to do all of the following to help us incorporate your contribution
quickly and easily:
- [ ] Make sure the PR title is formatted like:
`[BEAM-<Jira issue #>] Description of pull request`
- [ ] Make sure tests pass via `mvn clean verify`. (Even better, enable
Travis-CI on your fork and ensure the whole test matrix passes).
- [ ] Replace `<Jira issue #>` in the title with the actual Jira issue
number, if there is one.
- [ ] If this contribution is large, please file an Apache
[Individual Contributor License
Agreement](https://www.apache.org/licenses/icla.pdf).
---
Apply a no-op ParDo to any PTransform that is not consumed.
You can merge this pull request into a Git repository by running:
$ git pull https://github.com/tgroh/beam no_orphaned_dataflow_reads
Alternatively you can review and apply these changes as the patch at:
https://github.com/apache/beam/pull/2587.patch
To close this pull request, make a commit to your master/trunk branch
with (at least) the following in the commit message:
This closes #2587
----
commit d3e894f588d6991e1fd095c3ea5c79760dad7066
Author: Thomas Groh <[email protected]>
Date: 2017-04-19T00:25:59Z
Ensure all Read outputs are consumed in Dataflow
Apply a no-op ParDo to any PTransform that is not consumed.
----
> DataflowRunner drops Reads with no consumers
> --------------------------------------------
>
> Key: BEAM-2007
> URL: https://issues.apache.org/jira/browse/BEAM-2007
> Project: Beam
> Issue Type: Bug
> Components: runner-dataflow
> Reporter: Daniel Halperin
> Assignee: Thomas Groh
> Fix For: First stable release
>
>
> Basically, if a pipeline has "just" a Read with no consumers, the optimizer
> in Dataflow will drop it. To preserve Beam semantics, we do want to run the
> Read and drop its output, e.g., because the Read may have side effects that
> we're testing for.
> Is it possible with pipeline surgery to find such Reads and add an Identity
> ParDo to them?
--
This message was sent by Atlassian JIRA
(v6.3.15#6346)
