opencensus-jaeger-python

Opencensus Jaeger Exporter for Python

This is an example Python app using the Opencensus Jaeger exporter. It calculates a Fast Fourier Transform and annotates the parent span with the result. The code is in fft.py

Setup and Usage

To install the required packages, opencensus and thrift, using pip.

pip install opencensus thrift

Include these imports to use the tracer:

from opencensus.trace import execution_context
from opencensus.trace import tracer as tracer_module
from opencensus.trace.exporters import jaeger_exporter

Configure a tracer

exporter = jaeger_exporter.JaegerExporter(
    service_name="",
    host_name="",
    port="",
    username="auth",
    password=auth_token,
    endpoint="/v1/trace"
    )

tracer = tracer_module.Tracer(exporter=exporter)

Now the tracer can be saved so that it can be called from other functions.

tracer.store_tracer()

Retrieve the tracer from the execution context:

tracer = execution_context.get_opencensus_tracer()

Create Spans

To create a span, simply call tracer.span(name=span_name). Starting and closing a span can be simplified by using Python's with statement:

with tracer.span(name=span_name) as span:
    some_function()

# the span is automatically closed and exported

Alternatively, the span can be manually managed.

span = tracer.span(name=span_name)
span.start()
do_some_work()
span.finish()

The tracer is able to automatically manage parent/child spans if a child span is started while a parent span is still active.

with tracer.span(name="parent_span") as parent_span:
    with tracer.span(name="child_span") as child_span:
        do_some_work()

child_span will be exported as a child span of parent_span

A child span can also be manually created.

child_span = parent_span.span(name="child_span")

Running the example

Install the pip packages from the setup section.

Then run the script:

./fft.py

Resources

• https://opencensus.io/api/python/trace/usage.html