make force_flush() in PushMetricExporter synchronous

[gruebel]

Changes

• follow-up to leverage native async trait in PushMetricExporter #2662 (comment)

Merge requirement checklist

• CONTRIBUTING guidelines followed
• Unit tests added/updated (if applicable)
• Appropriate CHANGELOG.md files updated for non-trivial, user-facing changes
• Changes in public API reviewed (if applicable)

[codecov]

Codecov Report

Attention: Patch coverage is 0% with 5 lines in your changes missing coverage. Please review.

Project coverage is 79.1%. Comparing base (dbd44a3) to head (b201972).
Report is 1 commits behind head on main.

Files with missing lines	Patch %	Lines
opentelemetry-sdk/src/metrics/periodic_reader.rs	0.0%	2 Missing ⚠️
opentelemetry-otlp/src/metric.rs	0.0%	1 Missing ⚠️
...pentelemetry-sdk/src/metrics/in_memory_exporter.rs	0.0%	1 Missing ⚠️
opentelemetry-stdout/src/metrics/exporter.rs	0.0%	1 Missing ⚠️

Additional details and impacted files

@@          Coverage Diff          @@
##            main   #2663   +/-   ##
=====================================
  Coverage   79.1%   79.1%           
=====================================
  Files        122     122           
  Lines      22560   22560           
=====================================
  Hits       17858   17858           
  Misses      4702    4702           

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

[cijothomas]

LGTM, the changelog need to be moved.

[lalitb]

While this change is in right direction, and only affect the custom exporters, I am still thinking about the breaking changes with deprecating, if possible (specifically that we are 0.28 was RC), and concerns coming from #2645. Though it won't be trivial in this case, as we want to eventually keep the function name same. Which means, we need to deprecate the existing and bring new sync method under feature flag. Just for discussions - not to block this PR.

[cijothomas]

specifically that we are 0.28 was RC

Good point! Same issue in #2654 too.
Here's my thinking:
RC has much higher bar to make a breaking change. In this PR and the one linked above, I consider this is okay to do. The disruption is very minimal, so it is not worth the effort to do this in back-compat way (if there is a way!)

(Things would have been different if 1.0 was released - we absolutely cannot make this change or the one from #2654 post 1.0.)