Docs » Supported integrations in Splunk Observability Cloud » Instrument back-end applications to send spans to Splunk APM » Instrument Go applications for Splunk Observability Cloud » Connect Go trace data with logs for Splunk Observability Cloud

Connect Go trace data with logs for Splunk Observability Cloud πŸ”—

You can configure logging libraries to include tracing attributes provided automatically by Splunk OTel Go instrumentation. Use the trace metadata to correlate traces with log events and explore logs in Splunk.

The following steps describe how to extract trace metadata and configure logging libraries to include the data as log fields.

Extract trace metadata from contexts πŸ”—

The SpanContextFromContext function of the OpenTelemetry Trace API extracts the trace metadata from a context.Context and returns it in the form of a SpanContext. The following example shows how to use the function:

spanContext := trace.SpanContextFromContext(ctx)
if !spanContext.IsValid() {
        // ctx does not contain a valid span.
        // There is no trace metadata to add.
        return
}

The SpanContext struct contains the trace and span ID, trace flags with sampling information, and state information. You can add this metadata to log events to enrich their context, and to correlate traces and logs.

Annotate log events πŸ”—

After you gather metadata in a SpanContext, you can use it to annotate log events.

Structured logs πŸ”—

If you’re using a structured logger, add the trace metadata as fields. The following example shows how to annotate log events using the zap logging library:

logger, _ := zap.NewProduction()
defer logger.Sync()
logger = logger.With(
        zap.String("trace_id", spanContext.TraceID().String()),
        zap.String("span_id", spanContext.SpanID().String()),
        zap.String("trace_flags", spanContext.TraceFlags().String()),
)
logger.Info("Failed to fetch URL", zap.String("URL", url))

Unstructured logs πŸ”—

If you’re using unstructured logging, you can add the trace metadata as part of the logged message. The following example shows how to add trace metadata using the standard library log package:

// Add the metadata following an order you can parse later on
log.Printf(
        "(trace_id: %s, span_id: %s, trace_flags: %s): failed to fetch URL: %s",
        spanContext.TraceID().String(),
        spanContext.SpanID().String(),
        spanContext.TraceFlags().String(),
        url,
)

Example of log annotation πŸ”—

The following is a complete example of a chi server that extracts trace metadata and annotates log messages for handled requests:

package main

import (
        "context"
        "net/http"

        "github.com/go-chi/chi"
        "github.com/signalfx/splunk-otel-go/instrumentation/github.com/go-chi/chi/splunkchi"
        "go.opentelemetry.io/otel/trace"
        "go.uber.org/zap"
)

func withTraceMetadata(ctx context.Context, logger *zap.Logger) *zap.Logger {
        spanContext := trace.SpanContextFromContext(ctx)
        if !spanContext.IsValid() {
                // ctx does not contain a valid span.
                // There is no trace metadata to add.
                return logger
        }
        return logger.With(
                zap.String("trace_id", spanContext.TraceID().String()),
                zap.String("span_id", spanContext.SpanID().String()),
                zap.String("trace_flags", spanContext.TraceFlags().String()),
        )
}

func helloHandler(logger *zap.Logger) http.HandlerFunc {
        return func(w http.ResponseWriter, r *http.Request) {
                l := withTraceMetadata(r.Context(), logger)

                n, err := w.Write([]byte("Hello World!\n"))
                if err != nil {
                        w.WriteHeader(http.StatusInternalServerError)
                        l.Error("failed to write request response", zap.Error(err))
                } else {
                        l.Info("request handled", zap.Int("response_bytes", n))
                }
        }
}

func main() {
        logger, err := zap.NewProduction()
        if err != nil {
                panic(err)
        }
        defer logger.Sync()

        router := chi.NewRouter()
        router.Use(splunkchi.Middleware())
        router.Get("/hello", helloHandler(logger))
        if err := http.ListenAndServe(":8080", router); err != nil {
                panic(err)
        }
}