OpenTelemetry instrumentation for pino
This module provides automatic instrumentation of the pino
module to inject trace-context into Pino log records (log correlation) and to send Pino logging to the OpenTelemetry Logging SDK (log sending). It may be loaded using the @opentelemetry/sdk-trace-node
package and is included in the @opentelemetry/auto-instrumentations-node
bundle.
If total installation size is not constrained, it is recommended to use the @opentelemetry/auto-instrumentations-node
bundle with @opentelemetry/sdk-node for the most seamless instrumentation experience.
Compatible with OpenTelemetry JS API and SDK 1.0+
.
Installation
npm install --save @opentelemetry/instrumentation-pino
Supported Versions
pino
versions >=5.14.0 <10
- The "log sending" feature is only supported in pino v7 and later.
Usage
const { NodeSDK, tracing, logs, api } = require('@opentelemetry/sdk-node');
const { PinoInstrumentation } = require('@opentelemetry/instrumentation-pino');
const sdk = new NodeSDK({
spanProcessor: new tracing.SimpleSpanProcessor(new tracing.ConsoleSpanExporter()),
logRecordProcessor: new logs.SimpleLogRecordProcessor(new logs.ConsoleLogRecordExporter()),
instrumentations: [
new PinoInstrumentation({
}),
]
})
sdk.start();
const pino = require('pino');
const logger = pino();
logger.info('hi');
const tracer = api.trace.getTracer('example');
tracer.startActiveSpan('manual-span', span => {
logger.info('in a span');
});
Log sending
Creation of a Pino Logger will automatically add a Pino destination that sends log records to the OpenTelemetry Logs SDK. The OpenTelemetry SDK can be configured to handle those records -- for example, sending them on to an OpenTelemetry collector for log archiving and processing. The example above shows a minimal configuration that emits OpenTelemetry log records to the console for debugging.
If the OpenTelemetry SDK is not configured with a Logger provider, then this added destination will be a no-op.
Log sending can be disabled with the disableLogSending: true
option.
Log correlation
Pino logger calls in the context of a tracing span will have fields identifying the span added to the log record. This allows correlating log records with tracing data. The added fields are (spec):
trace_id
span_id
trace_flags
These field names can optionally be configured via the logKeys
option. For example:
new PinoInstrumentation({
logKeys: {
traceId: 'myTraceId',
spanId: 'mySpanId',
traceFlags: 'myTraceFlags',
},
}),
After adding these fields, the optional logHook
is called to allow injecting additional fields. For example:
logHook: (span, record) => {
record['resource.service.name'] = provider.resource.attributes['service.name'];
}
When no span context is active or the span context is invalid, injection is skipped.
Log injection can be disabled with the disableLogCorrelation: true
option.
Pino instrumentation options
Option | Type | Description |
---|
disableLogSending | boolean | Whether to disable log sending. Default false . |
disableLogCorrelation | boolean | Whether to disable log correlation. Default false . |
logKeys | record | A record with keys traceId , spanId , and traceFlags string fields giving the field names to use for log-correlation span data. |
logHook | LogHookFunction | An option hook to inject additional context to a log record after trace-context has been added. This requires disableLogCorrelation to be false. |
Semantic Conventions
This package does not currently generate any attributes from semantic conventions.
Alternative log sending with pino-opentelemetry-transport
A possible alternative to the "log sending" support provided by this instrumentation is the pino-opentelemetry-transport
package.
Every Pino logger has an output "destination", for example, a file or stdout. Since v7, Pino includes support for "transports", a type of destination that uses a worker thread to run the transport code. When calling logger.info("hi")
, Pino serializes the log record to a JSON string, sends that string to the worker for it to be handled.
The pino-opentelemetry-transport
package uses this mechanism. It starts an OpenTelemetry SDK LoggerProvider
in the worker thread, parses each log record string, translates it into the OpenTelemetry Logs data model and sends it. Note that this LoggerProvider
is independent of any OpenTelemetry SDK components in the main thread.
The log sending support in this instrumentation works on the main thread and uses the OpenTelemetry SDK configured in the main thread. Otherwise the two mechanisms are very similar. Note that because they are maintained separately, there might be small differences in how Pino log records are translated into the OpenTelemetry Logs data model.
Useful links
License
Apache 2.0 - See LICENSE for more information.