Skip to main content

OpenTelemetry span exporter that writes to stdout in OTLP format

Project description

OTLP Stdout Span Exporter for Python

PyPI version

A Python span exporter that writes OpenTelemetry spans to stdout, using a custom serialization format that embeds the spans serialized as OTLP protobuf in the payload field. The message envelope carries metadata about the spans, such as the service name, the OTLP endpoint, and the HTTP method:

{
  "__otel_otlp_stdout": "0.1.0",
  "source": "my-service",
  "endpoint": "http://localhost:4318/v1/traces",
  "method": "POST",
  "content-type": "application/x-protobuf",
  "content-encoding": "gzip",
  "headers": {
    "tenant-id": "tenant-12345",
    "custom-header": "value"
  },
  "payload": "<base64-encoded-gzipped-protobuf>",
  "base64": true,
  "level": "INFO"
}

Outputting telemetry data in this format directly to stdout makes the library easily usable in network constrained environments, or in environments that are particularly sensitive to the overhead of HTTP connections, such as AWS Lambda.

[!IMPORTANT] This package is part of the serverless-otlp-forwarder project and is designed for AWS Lambda environments. While it can be used in other contexts, it's primarily tested with AWS Lambda.

Features

  • Uses OTLP Protobuf serialization for efficient encoding
  • Applies GZIP compression with configurable levels
  • Detects service name from environment variables
  • Supports custom headers via environment variables
  • Supports log level for filtering in log aggregation systems
  • Supports writing to stdout or named pipe
  • Consistent JSON output format
  • Zero external HTTP dependencies
  • Lightweight and fast

Installation

pip install otlp-stdout-span-exporter

Usage

The recommended way to use this exporter is with the standard OpenTelemetry BatchSpanProcessor, which provides better performance by buffering and exporting spans in batches, or, in conjunction with the lambda-otel-lite package, with the LambdaSpanProcessor, which is particularly optimized for AWS Lambda.

You can create a simple tracer provider with the BatchSpanProcessor and the OTLPStdoutSpanExporter:

from opentelemetry import trace
from opentelemetry.sdk.trace import TracerProvider
from opentelemetry.sdk.trace.export import BatchSpanProcessor
from otlp_stdout_span_exporter import OTLPStdoutSpanExporter
from otlp_stdout_span_exporter.constants import LogLevel, OutputType

# Create and set the tracer provider
provider = TracerProvider()
trace.set_tracer_provider(provider)

# Create and register the exporter with default options (stdout output)
exporter = OTLPStdoutSpanExporter(gzip_level=6)

# Or with log level for filtering
debug_exporter = OTLPStdoutSpanExporter(
    gzip_level=6,
    log_level=LogLevel.DEBUG
)

# Or with named pipe output
pipe_exporter = OTLPStdoutSpanExporter(
    output_type=OutputType.PIPE  # Will write to /tmp/otlp-stdout-span-exporter.pipe
)

provider.add_span_processor(BatchSpanProcessor(exporter))

# Your instrumentation code here
tracer = trace.get_tracer(__name__)
with tracer.start_as_current_span("my-operation") as span:
    span.set_attribute("my.attribute", "value")

Configuration

Constructor Options

OTLPStdoutSpanExporter(
    # GZIP compression level (0-9, where 0 is no compression and 9 is maximum compression)
    # Will be overridden by environment variable if set
    gzip_level=6,
    
    # Log level for filtering in log aggregation systems
    # If not specified, no level field will be included in the output
    log_level=LogLevel.INFO,
    
    # Output type (stdout or pipe)
    # Defaults to OutputType.STDOUT if not specified
    output_type=OutputType.STDOUT
)

Environment Variables

The exporter respects the following environment variables:

  • OTEL_SERVICE_NAME: Service name to use in output
  • AWS_LAMBDA_FUNCTION_NAME: Fallback service name (if OTEL_SERVICE_NAME not set)
  • OTEL_EXPORTER_OTLP_HEADERS: Headers for OTLP export, used in the headers field
  • OTEL_EXPORTER_OTLP_TRACES_HEADERS: Trace-specific headers (which take precedence if conflicting with OTEL_EXPORTER_OTLP_HEADERS)
  • OTLP_STDOUT_SPAN_EXPORTER_COMPRESSION_LEVEL: GZIP compression level (0-9). Defaults to 6.
  • OTLP_STDOUT_SPAN_EXPORTER_LOG_LEVEL: Log level for filtering (debug, info, warn, error). If set, adds a level field to the output.
  • OTLP_STDOUT_SPAN_EXPORTER_OUTPUT_TYPE: Output type ("pipe" or "stdout"). Defaults to "stdout". If set to "pipe", writes to /tmp/otlp-stdout-span-exporter.pipe.

[!IMPORTANT] Environment variables always take precedence over constructor parameters. If both are specified, the environment variable value will be used.

[!NOTE] For security best practices, avoid including authentication credentials or sensitive information in headers. The serverless-otlp-forwarder infrastructure is designed to handle authentication at the destination, rather than embedding credentials in your telemetry data.

Output Format

The exporter writes JSON objects to stdout with the following structure:

{
  "__otel_otlp_stdout": "0.1.0",
  "source": "my-service",
  "endpoint": "http://localhost:4318/v1/traces",
  "method": "POST",
  "content-type": "application/x-protobuf",
  "content-encoding": "gzip",
  "headers": {
    "tenant-id": "tenant-12345",
    "custom-header": "value"
  },
  "base64": true,
  "payload": "<base64-encoded-gzipped-protobuf>",
  "level": "INFO"
}
  • __otel_otlp_stdout is a marker to identify the output of this exporter.
  • source is the emitting service name.
  • endpoint is the OTLP endpoint (defaults to http://localhost:4318/v1/traces and just indicates the signal type. The actual endpoint is determined by the process that forwards the data).
  • method is the HTTP method (always POST).
  • content-type is the content type (always application/x-protobuf).
  • content-encoding is the content encoding (always gzip).
  • headers is the headers defined in the OTEL_EXPORTER_OTLP_HEADERS and OTEL_EXPORTER_OTLP_TRACES_HEADERS environment variables.
  • payload is the base64-encoded, gzipped, Protobuf-serialized span data in OTLP format.
  • base64 is a boolean flag to indicate if the payload is base64-encoded (always true).
  • level is the log level (only present if configured via constructor or environment variable).

Named Pipe Output

When configured to use named pipe output (either via constructor or environment variable), the exporter will write to /tmp/otlp-stdout-span-exporter.pipe instead of stdout. This can be useful in environments where you want to process the telemetry data with a separate process.

If the pipe doesn't exist or can't be written to, the exporter will automatically fall back to stdout with a warning.

License

MIT

See Also

  • GitHub - The main project repository for the Serverless OTLP Forwarder project
  • GitHub | npm - The Node.js version of this exporter
  • GitHub | crates.io - The Rust version of this exporter

Project details


Download files

Download the file for your platform. If you're not sure which to choose, learn more about installing packages.

Source Distribution

otlp_stdout_span_exporter-0.16.0.tar.gz (8.5 kB view details)

Uploaded Source

Built Distribution

If you're not sure about the file name format, learn more about wheel file names.

otlp_stdout_span_exporter-0.16.0-py3-none-any.whl (10.1 kB view details)

Uploaded Python 3

File details

Details for the file otlp_stdout_span_exporter-0.16.0.tar.gz.

File metadata

File hashes

Hashes for otlp_stdout_span_exporter-0.16.0.tar.gz
Algorithm Hash digest
SHA256 0b39a3cca99a5beee9d2eaa763b6e3c0942e6c90904e9c31729b5659b0887037
MD5 1e98364480df0908f7ecb7767ef57afc
BLAKE2b-256 65fe67c608009832fcb28638a36753dfa48602f43310fedbdf5e150f9670c2cf

See more details on using hashes here.

File details

Details for the file otlp_stdout_span_exporter-0.16.0-py3-none-any.whl.

File metadata

File hashes

Hashes for otlp_stdout_span_exporter-0.16.0-py3-none-any.whl
Algorithm Hash digest
SHA256 bd7224d0823ef7c6ffda3a25cd9caa34392ba93677246cce4de2a287a559bb3d
MD5 8ebd5516238408b7282381fff1958ccc
BLAKE2b-256 f647f6a028c6ef71b9aa43e9e51fc6430268abea71beb736d0bbd19cb0fed436

See more details on using hashes here.

Supported by

AWS Cloud computing and Security Sponsor Datadog Monitoring Depot Continuous Integration Fastly CDN Google Download Analytics Pingdom Monitoring Sentry Error logging StatusPage Status page