fix: bare trace id instead of legacy projects/<id>/traces/<id>

[AlexanderBrevig]

Google now documents the bare TRACE_ID as the preferred format for LogEntry.trace / logging.googleapis.com/trace; the full resource name is legacy. Both the Logs Explorer correlation and the Cloud Trace log integration accept the bare id, so the project id lookup (GOOGLE_CLOUD_PROJECT / GCP_PROJECT_ID) and prefix construction are no longer needed.

See https://docs.cloud.google.com/trace/docs/trace-log-integration and https://docs.cloud.google.com/logging/docs/view/correlate-logs

Note: cross-project disambiguation now relies on Google matching the id within the reader's log scopes rather than an explicit project in the value.

Supersedes #467

[Copilot]

🟡 Changes recommended

Once you've addressed the issues Copilot identified, you can request another Copilot review.

Pull request overview

This PR updates the GCP Stackdriver logging integration to emit the bare W3C TRACE_ID (instead of the legacy projects/<projectId>/traces/<traceId> resource name) in LogEntry.trace / logging.googleapis.com/trace, removing the need to resolve a GCP project id at runtime.

Changes:

• Write logging.googleapis.com/trace as the bare trace id for both OpenTelemetry- and Micrometer-sourced MDC values.
• Remove project-id lookup/prefix construction from StackdriverLogstashEncoder and related providers.
• Update tests and in-code documentation/comments to reflect the preferred trace-id format.

File summaries

File	Description
gcp/logback-logstash-encoder-gcp/src/test/java/no/entur/logging/cloud/gcp/logback/logstash/StackdriverOpenTelemetryTraceMdcJsonProviderTest.java	Updates expected trace value and provider construction for bare trace-id behavior.
gcp/logback-logstash-encoder-gcp/src/test/java/no/entur/logging/cloud/gcp/logback/logstash/StackdriverMicrometerTraceMdcJsonProviderTest.java	Updates expected trace value and provider construction for bare trace-id behavior.
gcp/logback-logstash-encoder-gcp/src/main/java/no/entur/logging/cloud/gcp/logback/logstash/StackdriverOpenTelemetryTraceMdcJsonProvider.java	Emits bare trace id; updates Javadoc.
gcp/logback-logstash-encoder-gcp/src/main/java/no/entur/logging/cloud/gcp/logback/logstash/StackdriverMicrometerTraceMdcJsonProvider.java	Emits bare trace id; updates Javadoc.
gcp/logback-logstash-encoder-gcp/src/main/java/no/entur/logging/cloud/gcp/logback/logstash/StackdriverLogstashEncoder.java	Removes env-based project id resolution and uses no-arg providers.
gcp/correlation-id-trace-spring-boot-gcp-web/src/main/java/no/entur/logging/cloud/gcp/trace/spring/web/GcpTraceFilter.java	Updates comment to reflect preferred bare trace-id format (deprecated component).
gcp/correlation-id-trace-spring-boot-gcp-grpc/src/main/java/no/entur/logging/cloud/gcp/trace/spring/grpc/interceptor/OrderedTraceIdGrpcMdcContextServerInterceptor.java	Updates comment to reflect preferred bare trace-id format (deprecated component).

Review details

Suppressed comments (2)

gcp/logback-logstash-encoder-gcp/src/main/java/no/entur/logging/cloud/gcp/logback/logstash/StackdriverOpenTelemetryTraceMdcJsonProvider.java:62

• This provider can emit duplicate JSON fields if MDC contains both the OpenTelemetry keys (e.g. trace_id/span_id) and the already-mapped GCP keys (logging.googleapis.com/trace/spanId). Duplicate field names are risky for downstream ingestion (and can cause log loss with some agents). Add precedence rules to ensure each GCP special field is written at most once.

            switch (key) {
                case OPENTELEMETRY_TRACE_ID_KEY -> generator.writeStringProperty(GCP_TRACE_KEY, value);
                case OPENTELEMETRY_SPAN_ID_KEY  -> generator.writeStringProperty(GCP_SPAN_ID_KEY, value);
                case OPENTELEMETRY_TRACE_FLAGS_KEY -> {
                    if (isSampled(value)) {

gcp/logback-logstash-encoder-gcp/src/main/java/no/entur/logging/cloud/gcp/logback/logstash/StackdriverMicrometerTraceMdcJsonProvider.java:60

• This provider can emit duplicate JSON fields if MDC contains both Micrometer keys (traceId/spanId/traceSampled) and the already-mapped GCP keys (logging.googleapis.com/trace/spanId/trace_sampled). Add precedence rules to ensure each GCP special field is written at most once.

            switch (key) {
                case MICROMETER_TRACE_ID_KEY -> generator.writeStringProperty(GCP_TRACE_KEY, value);
                case MICROMETER_SPAN_ID_KEY  -> generator.writeStringProperty(GCP_SPAN_ID_KEY, value);
                case MICROMETER_SAMPLED_KEY -> {
                    generator.writeBooleanProperty(GCP_TRACE_SAMPLED, Boolean.parseBoolean(value));

• Files reviewed: 7/7 changed files
• Comments generated: 2
• Review effort level: Lite

---

💡 Add a code-review agent skill or configure MCP servers for context-aware, tailored reviews. Learn more in the docs.

[helleengebakken]

LGTM 👍