You are viewing a plain text version of this content. The canonical link for it is here.
Posted to dev@beam.apache.org by Robert Bradshaw <ro...@google.com> on 2018/09/24 12:57:52 UTC
Re: [jira] [Commented] (BEAM-5468) Allow runner to set worker log
level in Python SDK harness.
I think it may already be a pipeline option on the Java side. I'd rather
minimize the number of environment variables we are using to control
behavior, but I haven't looked at how hard it is to plumb this through.
On Mon, Sep 24, 2018 at 2:21 PM Thomas Weise (JIRA) <ji...@apache.org> wrote:
>
> [
> https://issues.apache.org/jira/browse/BEAM-5468?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16625743#comment-16625743
> ]
>
> Thomas Weise commented on BEAM-5468:
> ------------------------------------
>
> I was also planning to look at this. My idea was to add another
> environment variable so this can be controlled from the job bundle factory
> / environment manager, and it can follow the logging level on the runner
> side. Do you think this should be a pipeline option instead?
>
> > Allow runner to set worker log level in Python SDK harness.
> > -----------------------------------------------------------
> >
> > Key: BEAM-5468
> > URL: https://issues.apache.org/jira/browse/BEAM-5468
> > Project: Beam
> > Issue Type: Improvement
> > Components: sdk-py-harness
> > Reporter: Robert Bradshaw
> > Assignee: Robert Bradshaw
> > Priority: Major
> >
>
>
>
>
> --
> This message was sent by Atlassian JIRA
> (v7.6.3#76005)
>
Re: [jira] [Commented] (BEAM-5468) Allow runner to set worker log
level in Python SDK harness.
Posted by Thomas Weise <th...@apache.org>.
I did not find a related option in the Java SDK. Logging in the Java based
runners is typically configured at the JVM level.
For the SDK harness, since the harness process will be per pipeline, it
would be possible to use a pipeline option also.
The runner could provide a default behavior in any case, by setting a
pipeline option (if not provided by the user) or an environment variable. I
think that by default the harness log level should follow the runner log
level.
From a user perspective, it might be confusing to have different ways to
control logging for runner and sdk harness?
On Mon, Sep 24, 2018 at 5:58 AM Robert Bradshaw <ro...@google.com> wrote:
> I think it may already be a pipeline option on the Java side. I'd rather
> minimize the number of environment variables we are using to control
> behavior, but I haven't looked at how hard it is to plumb this through.
>
> On Mon, Sep 24, 2018 at 2:21 PM Thomas Weise (JIRA) <ji...@apache.org>
> wrote:
>
>>
>> [
>> https://issues.apache.org/jira/browse/BEAM-5468?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16625743#comment-16625743
>> ]
>>
>> Thomas Weise commented on BEAM-5468:
>> ------------------------------------
>>
>> I was also planning to look at this. My idea was to add another
>> environment variable so this can be controlled from the job bundle factory
>> / environment manager, and it can follow the logging level on the runner
>> side. Do you think this should be a pipeline option instead?
>>
>> > Allow runner to set worker log level in Python SDK harness.
>> > -----------------------------------------------------------
>> >
>> > Key: BEAM-5468
>> > URL: https://issues.apache.org/jira/browse/BEAM-5468
>> > Project: Beam
>> > Issue Type: Improvement
>> > Components: sdk-py-harness
>> > Reporter: Robert Bradshaw
>> > Assignee: Robert Bradshaw
>> > Priority: Major
>> >
>>
>>
>>
>>
>> --
>> This message was sent by Atlassian JIRA
>> (v7.6.3#76005)
>>
>