You are viewing a plain text version of this content. The canonical link for it is here.

Posted to issues@beam.apache.org by "Kenneth Knowles (Jira)" <ji...@apache.org> on 2022/06/03 23:05:00 UTC

[jira] [Commented] (BEAM-6708) Too many open files due grpc channel not being closed (direct runnner Pubsub source)

    [ https://issues.apache.org/jira/browse/BEAM-6708?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=17547499#comment-17547499 ] 

Kenneth Knowles commented on BEAM-6708:
---------------------------------------

This issue has been migrated to https://github.com/apache/beam/issues/19260

> Too many open files due grpc channel not being closed (direct runnner Pubsub source)
> ------------------------------------------------------------------------------------
>
>                 Key: BEAM-6708
>                 URL: https://issues.apache.org/jira/browse/BEAM-6708
>             Project: Beam
>          Issue Type: Bug
>          Components: runner-direct
>    Affects Versions: 2.10.0
>            Reporter: Stefan Boenders
>            Priority: P3
>
> When '_read_from_pubsub' (runnners/direct/transform_evaluator.py:448) is being invoked, it will create a Pubsub SubscriberClient. This will create a grpc channel which is not being closed. When you run a pipeline with the direct runner which has a pubsub source in it, it will run eventually into an error: "too many open files". This is because these grpc channels are never being closed.
> The suggestion is to wrap the code after the creation of the client in a try/finally block and close the channel (exposed on sub_client.api.transport._channel) in the finally block.
>  
> Some extra info about the issue: https://github.com/googleapis/google-cloud-python/issues/5523



--
This message was sent by Atlassian Jira
(v8.20.7#820007)