[
https://issues.apache.org/jira/browse/BEAM-10113?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=17548623#comment-17548623
]
Danny McCormick commented on BEAM-10113:
----------------------------------------
This issue has been migrated to https://github.com/apache/beam/issues/20246
> PubSubIO readMessagesWithMessageId() breaks the payload encoding when using
> DataflowRunner
> ------------------------------------------------------------------------------------------
>
> Key: BEAM-10113
> URL: https://issues.apache.org/jira/browse/BEAM-10113
> Project: Beam
> Issue Type: Bug
> Components: io-java-gcp
> Affects Versions: 2.20.0
> Environment: Running locally on MacOS Catalina 10.15.4 and as a
> Dataflow job in GCP.
> Reporter: Alexander Malyga
> Priority: P3
> Labels: GCP, dataflow, pubsubio, runner
> Fix For: Not applicable
>
>
> My pipeline reads PubSub messages and parses their payload to objects using
> Gson. I use PubsubIO.readMessagesWithMessageId() to get the PubSub message
> and the message ID.
> I tested the pipeline thoroughly by running it with the DirectRunner in my
> local machine and everything works fine, but when running it as a Dataflow
> job in GCP using the DataflowRunner, Gson can't parse the messages properly
> because the first character of the payload (opening bracket "{") is missing,
> this only happens when using the DataflowRunner.
> I noticed that the problem no longer happens when using
> PubsubIO.readStrings() instead of PubsubIO.readMessagesWithMessageId() and
> getting the payload directly (previously I had to decode the payload using
> new String(element.getPayload(), StandardCharsets.UTF_8); )
>
>
--
This message was sent by Atlassian Jira
(v8.20.7#820007)
