[ 
https://issues.apache.org/jira/browse/BEAM-2899?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16267774#comment-16267774
 ] 

ASF GitHub Bot commented on BEAM-2899:
--------------------------------------

kennknowles commented on issue #4105: [BEAM-2899]  Fork FnDataService from 
runners-core
URL: https://github.com/apache/beam/pull/4105#issuecomment-347361536
 
 
   LGTM

----------------------------------------------------------------
This is an automated message from the Apache Git Service.
To respond to the message, please log on GitHub and use the
URL above to go to the specific comment.
 
For queries about this service, please contact Infrastructure at:
us...@infra.apache.org


> Universal Local Runner
> ----------------------
>
>                 Key: BEAM-2899
>                 URL: https://issues.apache.org/jira/browse/BEAM-2899
>             Project: Beam
>          Issue Type: Improvement
>          Components: runner-core
>            Reporter: Henning Rohde
>            Assignee: Thomas Groh
>              Labels: portability
>
> To make the portability effort tractable, we should implement a Universal 
> Local Runner (ULR) in Java that runs in a single server process plus docker 
> containers for the SDK harness containers. It would serve multiple purposes:
>   (1) A reference implementation for other runners. Ideally, any new feature 
> should be implemented in the ULR first.
>   (2) A fully-featured test runner for SDKs who participate in the 
> portability framework. It thus complements the direct runners.
>   (3) A test runner for user code that depends on or customizes the runtime 
> environment. For example, a DoFn that shells out has a dependency that may be 
> satisfied on the user's desktop (and thus works fine on the direct runner), 
> but perhaps not by the container harness image. The ULR allows for an easy 
> way to find out.
> The Java direct runner presumably has lots of pieces that can be reused.



--
This message was sent by Atlassian JIRA
(v6.4.14#64029)

Reply via email to