[
https://issues.apache.org/jira/browse/BEAM-4826?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16566219#comment-16566219
]
Henning Rohde commented on BEAM-4826:
-------------------------------------
Yeah - unused outputs from ParDo is fine. As you say, they may occur from user
code naturally.
Flatten is the only transform I can think of, where the runner would split it
by input. We should just remove it in that case, like Dataflow does (because a
1 input flatten is a no-op) -- although for Go at least, it would work fine.
Other transforms generally have an expected input arity and the SDK can't
execute the bundle if the inputs are not all wired up. Mucking with such inputs
probably wouldn't work well.
> Flink runner sends bad flatten to SDK
> -------------------------------------
>
> Key: BEAM-4826
> URL: https://issues.apache.org/jira/browse/BEAM-4826
> Project: Beam
> Issue Type: Bug
> Components: runner-flink
> Reporter: Henning Rohde
> Assignee: Ankur Goenka
> Priority: Major
> Labels: portability
>
> For a Go flatten test w/ 3 input, the Flink runner splits this into 3 bundle
> descriptors. But it sends the original 3-input flatten but w/ 1 actual input
> present in each bundle descriptor. This is inconsistent and the SDK shouldn't
> expect dangling PCollections. In contrast, Dataflow removes the flatten when
> it does the same split.
> Snippet:
> register: <
> process_bundle_descriptor: <
> id: "3"
> transforms: <
> key: "e4"
> value: <
> unique_name: "github.com/apache/beam/sdks/go/pkg/beam.createFn'1"
> spec: <
> urn: "urn:beam:transform:pardo:v1"
> payload: [...]
> >
> inputs: <
> key: "i0"
> value: "n3"
> >
> outputs: <
> key: "i0"
> value: "n4"
> >
> >
> >
> transforms: <
> key: "e7"
> value: <
> unique_name: "Flatten"
> spec: <
> urn: "beam:transform:flatten:v1"
> >
> inputs: <
> key: "i0"
> value: "n2"
> >
> inputs: <
> key: "i1"
> value: "n4" . // <----------- only one present.
> >
> inputs: <
> key: "i2"
> value: "n6"
> >
> outputs: <
> key: "i0"
> value: "n7"
> >
> >
> >
> [...]
--
This message was sent by Atlassian JIRA
(v7.6.3#76005)
