Resource auto detection logging

[adamegyed]

Which problem is this PR solving?

• Fixes Add logging to Resource detection #920

Short description of the changes

• Added ResourceDetectionConfig interface to be optionally taken in by the detectResources function. Currently only configurable field is the logger, which must adhere to the logging interface specified in opentelemetry-api

• Added unit tests for logger functionality

[linux-foundation-easycla]

The committers are authorized under a signed CLA.

• ✅ Adam Egyed (07f0e0e, f0486e7)

[codecov]

Codecov Report

Merging #1211 into master will increase coverage by 0.30%.
The diff coverage is 96.55%.

@@            Coverage Diff             @@
##           master    #1211      +/-   ##
==========================================
+ Coverage   92.71%   93.01%   +0.30%     
==========================================
  Files         129      131       +2     
  Lines        3471     3696     +225     
  Branches      712      747      +35     
==========================================
+ Hits         3218     3438     +220     
- Misses        253      258       +5     

Impacted Files	Coverage Δ
...ry-resources/src/platform/node/detect-resources.ts	96.29% <94.44%> (-3.71%)	⬇️
...rces/src/platform/node/detectors/AwsEc2Detector.ts	87.50% <100.00%> (+0.40%)	⬆️
...sources/src/platform/node/detectors/EnvDetector.ts	95.55% <100.00%> (+7.46%)	⬆️
...sources/src/platform/node/detectors/GcpDetector.ts	95.34% <100.00%> (+0.22%)	⬆️
packages/opentelemetry-plugin-grpc/src/grpc.ts	96.79% <0.00%> (ø)
packages/opentelemetry-plugin-grpc/src/utils.ts	93.75% <0.00%> (ø)

[mwear]

I have one comment if we stick with the approach that is currently being taken.

An alternative approach that we might consider, is passing the ResourceDetectionConfig (with the logger) into each detector where they can make use it and add more detailed logging around events encountered internally. I'll let others weigh in on the merits of the different approaches.

[adamegyed]

@mwear I've modified the Detector interface to optionally take in the ResourceDetectionConfig, but I'm not familiar with what information is valuable from each detector. Perhaps another contributor could make use of the logger within the detectors.

[obecny]

Thank you for the changes, they looks good with one thing only you don't need to create a new NoopLogger in each detector as the logger will already be on config then - I added comment to each of this case otherwise you will always have NoopLogger instead of user logger.
The last thing is adding a tsdoc for all exported items

[mwear]

In addition to the typo below, @obecny was asking for tsdoc updates. I think he's looking for documentation about the config parameter added to the detect methods.

[obecny]

lgtm, thx for all the changes :)

[mwear]

Nice work! Thanks @adamegyed!