[
https://issues.apache.org/jira/browse/BEAM-10395?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]
Andrew Pilloud updated BEAM-10395:
----------------------------------
Resolution: Fixed
Status: Resolved (was: Open)
Fixed by https://github.com/apache/beam/pull/12144
> Dataflow runner should deduplicate files to stage by destination
> -----------------------------------------------------------------
>
> Key: BEAM-10395
> URL: https://issues.apache.org/jira/browse/BEAM-10395
> Project: Beam
> Issue Type: Improvement
> Components: runner-dataflow
> Reporter: Steve Niemitz
> Priority: P3
> Fix For: 2.24.0
>
> Time Spent: 1h 50m
> Remaining Estimate: 0h
>
> If a pipeline contains multiple files with the same destination path, the
> dataflow runner will try to stage them both in parallel, resulting in the
> upload usually failing (due to conflicting uploads).
> The runner should only upload one file per destination, and ideally check
> that the sources are the same as well.
--
This message was sent by Atlassian Jira
(v8.3.4#803005)
