You are viewing a plain text version of this content. The canonical link for it is here.
Posted to issues@beam.apache.org by "Mark Liu (JIRA)" <ji...@apache.org> on 2018/12/14 00:16:00 UTC

[jira] [Resolved] (BEAM-5058) Python precommits should run E2E tests

     [ https://issues.apache.org/jira/browse/BEAM-5058?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]

Mark Liu resolved BEAM-5058.
----------------------------
       Resolution: Done
    Fix Version/s: Not applicable

> Python precommits should run E2E tests
> --------------------------------------
>
>                 Key: BEAM-5058
>                 URL: https://issues.apache.org/jira/browse/BEAM-5058
>             Project: Beam
>          Issue Type: Task
>          Components: sdk-py-core, testing
>            Reporter: Udi Meiri
>            Assignee: Mark Liu
>            Priority: Major
>             Fix For: Not applicable
>
>          Time Spent: 6h 50m
>  Remaining Estimate: 0h
>
> According to [https://beam.apache.org/contribute/testing/] (which I'm working on), end-to-end tests should be run in precommit on each combination of \{batch, streaming}x\{SDK language}x\{supported runner}.
> At least 2 tests need to be added to Python's precommit: wordcount and wordcount_streaming on Dataflow, and possibly on other supported runners (direct runner and new runners plz).
>  These tests should be configured to run from a Gradle sub-project, so that they're run in parallel to the unit tests.
> Example that parallelizes Java precommit integration tests: [https://github.com/apache/beam/pull/5731]



--
This message was sent by Atlassian JIRA
(v7.6.3#76005)