You've already forked opentelemetry-go
mirror of
https://github.com/open-telemetry/opentelemetry-go.git
synced 2026-06-03 18:35:08 +02:00
Tyler Yahn
d03b03395d
Fix #7673 [Issue being addressed](https://github.com/open-telemetry/opentelemetry-go/issues/7673#issuecomment-3618325229): > 1. [`fn`](https://github.com/open-telemetry/opentelemetry-go/blob/1bc9713ac6dc8cbe2fd04fd6dc716d316059eb90/exporters/otlp/otlplog/otlploggrpc/internal/retry/retry_test.go#L163-L165) is [called](https://github.com/open-telemetry/opentelemetry-go/blob/1bc9713ac6dc8cbe2fd04fd6dc716d316059eb90/exporters/otlp/otlplog/otlploggrpc/internal/retry/retry.go#L87) > 2. It [returns an error](https://github.com/open-telemetry/opentelemetry-go/blob/1bc9713ac6dc8cbe2fd04fd6dc716d316059eb90/exporters/otlp/otlplog/otlploggrpc/internal/retry/retry_test.go#L165) > 3. The code [checks if the error is retryable](https://github.com/open-telemetry/opentelemetry-go/blob/1bc9713ac6dc8cbe2fd04fd6dc716d316059eb90/exporters/otlp/otlplog/otlploggrpc/internal/retry/retry.go#L92), it [always is](https://github.com/open-telemetry/opentelemetry-go/blob/1bc9713ac6dc8cbe2fd04fd6dc716d316059eb90/exporters/otlp/otlplog/otlploggrpc/internal/retry/retry_test.go#L149) > 4. [Time delay is checked](https://github.com/open-telemetry/opentelemetry-go/blob/1bc9713ac6dc8cbe2fd04fd6dc716d316059eb90/exporters/otlp/otlplog/otlploggrpc/internal/retry/retry.go#L97-L108) > - [Max elsapsed time](https://github.com/open-telemetry/opentelemetry-go/blob/1bc9713ac6dc8cbe2fd04fd6dc716d316059eb90/exporters/otlp/otlplog/otlploggrpc/internal/retry/retry_test.go#L156-L157) is 10 ms > - Initial [delay is 1ms](https://github.com/open-telemetry/opentelemetry-go/blob/1bc9713ac6dc8cbe2fd04fd6dc716d316059eb90/exporters/otlp/otlplog/otlploggrpc/internal/retry/retry_test.go#L151) > - Delay is determined to be 1ms > - The program proceeds to waiting > 5. [Wait is called](https://github.com/open-telemetry/opentelemetry-go/blob/1bc9713ac6dc8cbe2fd04fd6dc716d316059eb90/exporters/otlp/otlplog/otlploggrpc/internal/retry/retry.go#L110-L112) > 6. The [wait select statement is evaluated](https://github.com/open-telemetry/opentelemetry-go/blob/1bc9713ac6dc8cbe2fd04fd6dc716d316059eb90/exporters/otlp/otlplog/otlploggrpc/internal/retry/retry.go#L127-L138) > - On slow systems both `case`s are true > - [Non-deterministically](https://go.dev/ref/spec#:~:text=If%20one%20or,communications%20can%20proceed.) the [timer channel `case`](https://github.com/open-telemetry/opentelemetry-go/blob/1bc9713ac6dc8cbe2fd04fd6dc716d316059eb90/exporters/otlp/otlplog/otlploggrpc/internal/retry/retry.go#L137) is selected > - The retry function is re-run and and second iteration is recorded causing the failure > - On fast systems only the context cancel is true > - The retry stops here with only `1` execution Do not rely on non-deterministic `select` statement to catch ended context prior to waiting for a retry delay. Explicitly check the context prior to entering the wait. This resolves the flaky test and ensure in normal operation that requests with canceled context are ended without having to wait for any additional delays.