Internal - delete unused benchmark.yml workflow

[cleverchuk]

Summary

Fix feature test failures caused by the upstream opentelemetry-agent-for-testing overriding the otel.propagators property, remove unused benchmark workflow, and fix minor code style issues.

Details

Fix propagator override in test agent

The upstream opentelemetry-agent-for-testing jar contains an AutoConfigurationCustomizerProvider that overrides otel.propagators set by SolarwindsPropertiesSupplier. This caused SolarwindsContextPropagator to be absent from the composite propagator during tests, resulting in 3 test failures (context propagation and trigger trace tests).

Since DefaultConfigProperties resolution order is systemProperties > environmentVariables > defaultProperties, explicitly setting -Dotel.propagators=tracecontext,baggage,solarwinds as a JVM system property in the feature-tests module guarantees the SolarWinds propagator is always registered regardless of supplier ordering.

Remove unused benchmark workflow

The benchmark.yml GitHub Actions workflow was unused and removed.

Code style fixes

• Renamed test method to use consistent casing (W3C → W3c)
• Renamed unused catch variable (ignore → ignored)

Test services data

1. e-1712644058766987264
2. e-1712643928659124224
3. e-1742334541200846848
4. e-1777406072376840192

[Copilot]

Pull request overview

Removes the standalone GitHub Actions workflow used to run benchmarks (Benchmark), with the intent to rely on the benchmark job that exists in the Push workflow instead.

Changes:

• Deleted the .github/workflows/benchmark.yml workflow definition.

Comments suppressed due to low confidence (1)

.github/workflows/benchmark.yml:1

• Deleting this workflow removes features that the PR description says are already replaced: this file provided a manually-triggered (workflow_dispatch) benchmark run with inputs (agent-type/container-logs) and persisted results by committing to the benchmark-results branch. The current .github/workflows/push.yml benchmark job (around lines 423-458) only runs ./gradlew test and uploads artifacts, and there is still documentation pointing users to run the Benchmark workflow and to fetch results from benchmark-results (see benchmark/README.md:8-30). Either update the push workflow/docs to match the previous behavior, or clarify in the PR description that the result-publishing/manual-run functionality is being intentionally removed.

[jerrytfleung]

LGTM

[jerrytfleung]

LGTM