Lower various logging statement levels to clean up example printing

[hjtran]

The playground examples have tons of info logging that drown out any actual logging that the example might've caused. This is somewhat of a big hit to the playground experience (not just noted by me, but by others that I've been introducing to Beam)

I attempted a fix to change the playground's log level but that didn't seem to take. Vlado Djerek also tried to figure out where the logging level is set but was also unsuccessful.

In lieu of a real fix, I just changed or deleted various logging statements that I suspect are probably not all that useful. I suspect especially that the translation prints which really clog up the examples don't offer much useful information.

Here's an example of what running the example looks like right now:

[hjtran]

reviewer: @tvalentyn

[hjtran]

I left a few statements out because they seemed like they're more important, but I think they do still distract from the examples. If we ever find where the log level is set, I wonder if we should just disable the logging altogether and update any examples that use the logger to just use beam.Map(print) or beam.LogElements instead

[github-actions]

Stopping reviewer notifications for this pull request: review requested by someone other than the bot, ceding control

[codecov]

Codecov Report

All modified and coverable lines are covered by tests ✅

Project coverage is 71.44%. Comparing base (827e96d) to head (4ed3742).
Report is 135 commits behind head on master.

Additional details and impacted files

@@             Coverage Diff             @@
##           master   #30782       +/-   ##
===========================================
+ Coverage   38.52%   71.44%   +32.91%     
===========================================
  Files         698      710       +12     
  Lines      102371   104812     +2441     
===========================================
+ Hits        39440    74878    +35438     
+ Misses      61301    28304    -32997     
  Partials     1630     1630               

Flag	Coverage Δ
python	81.22% <100.00%> (+52.07%)	⬆️

Flags with carried forward coverage won't be shown. Click here to find out more.

☔ View full report in Codecov by Sentry.
📢 Have feedback on the report? Share it here.

[tvalentyn]

I attempted a fix to change the playground's #29948 but that didn't seem to take. Vlado Djerek also tried to figure out where the logging level is set but was also unsuccessful.

By crashing playground intepreter, I see that the main python entrypoint is located in some generated files, like:
/opt/playground/backend/executable_files/8ff1f92b-37b3-4fd7-9a08-fb8d1768ae5f/8ff1f92b-37b3-4fd7-9a08-fb8d1768ae5f.py

We could try to find a place where that is generated and add a line like logging.getLogger().setLevel(logging.WARNING) at the top of the file.

[hjtran]

I attempted a fix to change the playground's #29948 but that didn't seem to take. Vlado Djerek also tried to figure out where the logging level is set but was also unsuccessful.

By crashing playground intepreter, I see that the main python entrypoint is located in some generated files, like: /opt/playground/backend/executable_files/8ff1f92b-37b3-4fd7-9a08-fb8d1768ae5f/8ff1f92b-37b3-4fd7-9a08-fb8d1768ae5f.py

We could try to find a place where that is generated and add a line like logging.getLogger().setLevel(logging.WARNING) at the top of the file.

I attempted this in another PR, though I don't actually know how to test it. I just removed the log level setting altogether, but we could change to just WARNING. IMO though, playground examples seem like an unlikely place to debug runner issues, and the examples are always easy to run locally anyways if we ever find something amiss.

[tvalentyn]

I just removed the log level setting altogether, but we could change to just WARNING.

+1 to change to warning instead of removing

[tvalentyn]

(closed accidentally)

[tvalentyn]

I also don't known how to test playground code. a log level change should be safe to merge.