Chaos Toolkit Extension for Open Tracing

This project is an extension for the Chaos Toolkit for OpenTracing and OpenTelemetry.

Here is an example of what it could look like with the Jaeger backend.

Install

This package requires Python 3.9+

To be used from your experiment, this package must be installed in the Python environment where chaostoolkit already lives.

$ pip install -U chaostoolkit-opentracing

Usage

This extension provides two controls to trace your Chaos Toolkit experiment:

• Open Telemetry
• Open Tracing (legacy)

The only supported one is Open Telemetry as the Open Tracing is no longer maintained.

Open Telemetry

To enable Open Telemetry tracing, simply add the following control to your experiment:

{
    "controls": [
      {
          "name": "opentelemetry",
          "provider": {
              "type": "python",
              "module": "chaostracing.oltp"
          }
      }
    ]
}

We suggest you make it the first extension so it runs before and after all other extensions.

To configure the various Open Telemetry settings, please use the standard OLTP environment variables:

• the sdk variables
• the exporter variables

Mostly, you should set:

• OTEL_EXPORTER_OTLP_TRACES_ENDPOINT to point to your collector (for instance: http://localhost:4318/v1/traces)
• OTEL_EXPORTER_OTLP_TRACES_HEADERS to set any headers to pass to the exporter

NOTE: This extension supports OLTP over HTTP but not gRPC.

You can also instrument a variety of frameworks like this:

{
    "controls": [
      {
          "name": "opentelemetry",
          "provider": {
              "type": "python",
              "module": "chaostracing.oltp",
              "arguments": {
                "trace_httpx": true,
                "trace_requests": true,
                "trace_botocore": true
              }
          }
      }
    ]
}

This will enable the according instrumentation automatically.

AWS

This extension supports AWS X-Ray directly. Simply set the following variable:

export OTEL_VENDOR=aws

This can also be set in the configuration block:

{
    "configuration": {
        "otel_vendor": "aws"
    }
}

Google Cloud Platform Traces

If you intend on using Google Cloud Platform to export your traces to, please consider also installing the followings:

$ pip install opentelemetry-exporter-gcp-trace \
    opentelemetry-resourcedetector-gcp \
    opentelemetry-propagator-gcp

To authenticate the client, you can either:

• set GOOGLE_APPLICATION_CREDENTIALS environment variable
• pass the otel_gcp_service_account and otel_gcp_project_id variables in the configuration block
• set the CHAOSTOOLKIT_OTEL_GCP_SA and CHAOSTOOLKIT_OTEL_GCP_PROJECT_ID environment variables

In all cases, point to a service account which has the roles/cloudtrace.agent role as nthe name of the target project.

Finally, set the following variable:

export OTEL_VENDOR=gcp

This can also be set in the configuration block:

{
    "configuration": {
        "otel_vendor": "gcp"
    }
}

Azure Traces

To use this package to send traces to Azure monitors, please install the dependencies as follows:

$ pip install chaostoolkit-opentracing[azure]

Then set the APPLICATIONINSIGHTS_CONNECTION_STRING environment variable appropriately.

Finally, set the following variable:

export OTEL_VENDOR=azure

This can also be set in the configuration block:

{
    "configuration": {
        "otel_vendor": "azure"
    }
}

See Azure documentation for more details:

• https://learn.microsoft.com/en-us/python/api/overview/azure/core-tracing-opentelemetry-readme
• https://learn.microsoft.com/en-us/python/api/overview/azure/monitor-opentelemetry-exporter-readme

Other Open Telemetry vendors

Other vendors should work out of the box with the default settings. Otherwise, feel free to open an issue.

Legacy Open Tracing

This extensions supports the Open Tracing export format but highly recommends you to switch to Open Telemetry instead. There will be no support for Open Tracing support.

NOTE: Please see at the bottom of the page all the supported clients and exporters this control supports.

Declare within the experiment

To use this control, you can declare it on a per experiment basis like this:

{
    "configuration": {
        "tracing_provider": "jaeger",
        "tracing_host": "127.0.0.1",
        "tracing_port": 6831,
        "tracing_propagation": "b3"
    },
    "controls": [
        {
            "name": "opentracing",
            "provider": {
                "type": "python",
                "module": "chaostracing.control"
            }
        }
    ]
}

This will automatically create a Jaeger client to emit traces onto the address 127.0.0.1:6831 (over UDP).

Declare within the settings

You may also declare the control to be applied to all experiments by declaring the control from within the Chaos Toolkit settings file. In that case, you do not need to set the configuration or the controls at the experiment level and the control will be applied to every experiments you run.

controls:
  opentracing:
    provider:
      type: python
      module: chaostracing.control
      arguments:
        provider: jaeger
        host: 127.0.0.1
        port: 6831
        propagation: b3

Send traces from other extensions

You may also access the tracer from other extensions as follows.

For instance, assuming you have an extension that makes a HTTP call you want to trace specifically, you could do this from your extension's code:

from chaoslib import Configuration, Secrets
import requests
import opentracing

def some_function(configuration: Configuration, secrets: Secrets):
    tracer = opentracing.global_tracer()
    scope = tracer.scope_manager.active
    parent = scope.span

    with tracer.start_span("call-service1", child_of=parent) as span:
        span.set_tag('http.method','GET')
        span.set_tag('http.url', url)
        span.set_tag('span.kind', 'client')
        span.tracer.inject(span, 'http_headers', headers)

        r = requests.get(url, headers=headers)
        span.set_tag('http.status_code', r.status_code)

Because the opentracing exposes a noop tracer when non has been initialized, it should be safe to have that code in your extensions without having to determine if the extension has been enabled in the experiment.

Please note that, Open Tracing scope cannot be shared across threads (while spans can). So, when running this in a background activity, the tracer will not actually be set to the one that was initialized.

Open Tracing Provider Support

Jaeger tracer

The Jager tracer relies on the OpenTracing protocol which has now be superseded by OpenTelemetry. However, we still provide support for it.

To install the necessary dependencies for the Jaeger tracer, please run:

$ pip install -U jaeger-client~=4.8

Use the following configuration:

{
    "configuration": {
        "tracing_provider": "jaeger",
        "tracing_host": "127.0.0.1",
        "tracing_port": 6831,
        "tracing_propagation": "b3"
    },
    "controls": [
        {
            "name": "opentracing",
            "provider": {
                "type": "python",
                "module": "chaostracing.control"
            }
        }
    ]
}

Test

To run the tests for the project execute the following:

$ pdm run test

Contribute

If you wish to contribute more functions to this package, you are more than welcome to do so. Please, fork this project, make your changes following the usual PEP 8 code style, sprinkling with tests and submit a PR for review.

The Chaos Toolkit projects require all contributors must sign a Developer Certificate of Origin on each commit they would like to merge into the master branch of the repository. Please, make sure you can abide by the rules of the DCO before submitting a PR.