Advanced Usage¶

Agent Configuration¶

If the Datadog Agent is on a separate host from your application, you can modify the default ddtrace.tracer object to utilize another hostname and port. Here is a small example showcasing this:

from ddtrace import tracer

tracer.configure(hostname=<YOUR_HOST>, port=<YOUR_PORT>, https=<True/False>)

By default, these will be set to localhost, 8126, and False respectively.

You can also use a Unix Domain Socket to connect to the agent:

from ddtrace import tracer

tracer.configure(uds_path="/path/to/socket")

Context¶

The ddtrace.context.Context object is used to represent the state of a trace at a point in time. This state includes the trace id, active span id, distributed sampling decision and more. It is used to propagate the trace across execution boundaries like processes (Distributed Tracing), threads and tasks.

To retrieve the context of the currently active trace use:

context = tracer.current_trace_context()

Note that if there is no active trace then None will be returned.

Tracing Context Management¶

In ddtrace “context management” is the management of which ddtrace.Span or ddtrace.context.Context is active in an execution (thread, task, etc). There can only be one active span or context per execution at a time.

Context management enables parenting to be done implicitly when creating new spans by using the active span as the parent of a new span. When an active span finishes its parent becomes the new active span.

tracer.trace() automatically creates new spans as the child of the active context:

# Here no span is active
assert tracer.current_span() is None

with tracer.trace("parent") as parent:
    # Here `parent` is active
    assert tracer.current_span() is parent

    with tracer.trace("child") as child:
        # Here `child` is active.
        # `child` automatically inherits from `parent`
        assert tracer.current_span() is child

    # `parent` is active again
    assert tracer.current_span() is parent

# Here no span is active again
assert tracer.current_span() is None

Important

Span objects are owned by the execution in which they are created and must be finished in the same execution. To continue a trace in a different execution then the span.context can be passed between executions and activated. See the sections below for how to propagate spans and traces across task, thread or process boundaries.

Tracing Across Threads¶

To continue a trace across threads the context needs to be passed between threads:

import threading, time
from ddtrace import tracer

def _target(trace_ctx):
    tracer.context_provider.activate(trace_ctx)
    with tracer.trace("second_thread") as span2:
        # span2's parent will be the main_thread span
        time.sleep(1)

with tracer.trace("main_thread") as span:
    thread = threading.Thread(target=_target, args=(span.context,))
    thread.start()
    thread.join()

Tracing Across Processes¶

Just like the threading case, if tracing across processes is desired then the span has to be propagated as a context:

from multiprocessing import Process
import time
from ddtrace import tracer

def _target(ctx):
    tracer.context_provider.activate(ctx)
    with tracer.trace("proc"):
        time.sleep(1)
    tracer.shutdown()

with tracer.trace("work") as span:
    proc = Process(target=_target, args=(span.context,))
    proc.start()
    time.sleep(1)
    proc.join()

fork¶

If using fork(), any open spans from the parent process must be finished by the parent process. Any active spans from the original process will be converted to contexts to avoid memory leaks.

Here’s an example of tracing some work done in a child process:

import os, sys, time
from ddtrace import tracer

span = tracer.trace("work")

pid = os.fork()

if pid == 0:
    with tracer.trace("child_work"):
        time.sleep(1)
    sys.exit(0)

# Do some other work in the parent
time.sleep(1)
span.finish()
_, status = os.waitpid(pid, 0)
exit_code = os.WEXITSTATUS(status)
assert exit_code == 0

Tracing Across Asyncio Tasks¶

By default the active context will by propagated across tasks on creation as the contextvars context is copied between tasks. If this is not desirable then None can be activated in the new task:

tracer.context_provider.activate(None)

Manual Management¶

Parenting can be managed manually by using tracer.start_span() which by default does not activate spans when they are created. See the documentation for ddtrace.Tracer.start_span().

Context Providers¶

The default context provider used in the tracer uses contextvars to store the active context per execution. This means that any asynchronous library that uses contextvars will have support for automatic context management.

If there is a case where the default is insufficient then a custom context provider can be used. It must implement the ddtrace.provider.BaseContextProvider interface and can be configured with:

tracer.configure(context_provider=MyContextProvider)

Distributed Tracing¶

To trace requests across hosts, the spans on the secondary hosts must be linked together by setting trace_id and parent_id.

On the server side, it means to read propagated attributes and set them to the active tracing context.
On the client side, it means to propagate the attributes, commonly as a header/metadata.

ddtrace already provides default propagators but you can also implement your own.

Web Frameworks¶

Some web framework integrations support distributed tracing out of the box.

Supported web frameworks:

Framework/Library	Enabled
aiohttp	True
Bottle	True
Django	True
Falcon	True
Flask	True
Pylons	True
Pyramid	True
Requests	True
Tornado	True

HTTP Client¶

For distributed tracing to work, necessary tracing information must be passed alongside a request as it flows through the system. When the request is handled on the other side, the metadata is retrieved and the trace can continue.

To propagate the tracing information, HTTP headers are used to transmit the required metadata to piece together the trace.

class ddtrace.propagation.http.HTTPPropagator¶

A HTTP Propagator using HTTP headers as carrier.

static inject(span_context: ddtrace.context.Context, headers: Dict[str, str]) → None¶

Inject Context attributes that have to be propagated as HTTP headers.

Here is an example using requests:

import requests
from ddtrace.propagation.http import HTTPPropagator

def parent_call():
    with tracer.trace('parent_span') as span:
        headers = {}
        HTTPPropagator.inject(span.context, headers)
        url = '<some RPC endpoint>'
        r = requests.get(url, headers=headers)

Parameters

span_context (Context) – Span context to propagate.
headers (dict) – HTTP headers to extend with tracing attributes.

static extract(headers: Dict[str, str]) → ddtrace.context.Context¶

Extract a Context from HTTP headers into a new Context.

Here is an example from a web endpoint:

from ddtrace.propagation.http import HTTPPropagator

def my_controller(url, headers):
    context = HTTPPropagator.extract(headers)
    if context:
        tracer.context_provider.activate(context)

    with tracer.trace('my_controller') as span:
        span.set_meta('http.url', url)

Parameters: headers (dict) – HTTP headers to extract tracing attributes.
Returns: New Context with propagated attributes.

Custom¶

You can manually propagate your tracing context over your RPC protocol. Here is an example assuming that you have rpc.call function that call a method and propagate a rpc_metadata dictionary over the wire:

# Implement your own context propagator
class MyRPCPropagator(object):
    def inject(self, span_context, rpc_metadata):
        rpc_metadata.update({
            'trace_id': span_context.trace_id,
            'span_id': span_context.span_id,
        })

    def extract(self, rpc_metadata):
        return Context(
            trace_id=rpc_metadata['trace_id'],
            span_id=rpc_metadata['span_id'],
        )

# On the parent side
def parent_rpc_call():
    with tracer.trace("parent_span") as span:
        rpc_metadata = {}
        propagator = MyRPCPropagator()
        propagator.inject(span.context, rpc_metadata)
        method = "<my rpc method>"
        rpc.call(method, metadata)

# On the child side
def child_rpc_call(method, rpc_metadata):
    propagator = MyRPCPropagator()
    context = propagator.extract(rpc_metadata)
    tracer.context_provider.activate(context)

    with tracer.trace("child_span") as span:
        span.set_meta('my_rpc_method', method)

Resolving deprecation warnings¶

Before upgrading, it’s a good idea to resolve any deprecation warnings raised by your project. These warnings must be fixed before upgrading, otherwise the ddtrace library will not work as expected. Our deprecation messages include the version where the behavior is altered or removed.

In Python, deprecation warnings are silenced by default. To enable them you may add the following flag or environment variable:

$ python -Wall app.py

# or

$ PYTHONWARNINGS=all python app.py

Trace Filtering¶

It is possible to filter or modify traces before they are sent to the Agent by configuring the tracer with a filters list. For instance, to filter out all traces of incoming requests to a specific url:

from ddtrace import tracer

tracer.configure(settings={
    'FILTERS': [
        FilterRequestsOnUrl(r'http://test\.example\.com'),
    ],
})

The filters in the filters list will be applied sequentially to each trace and the resulting trace will either be sent to the Agent or discarded.

Built-in filters

The library comes with a FilterRequestsOnUrl filter that can be used to filter out incoming requests to specific urls:

class ddtrace.filters.FilterRequestsOnUrl(regexps)¶

Filter out traces from incoming http requests based on the request’s url.

This class takes as argument a list of regular expression patterns representing the urls to be excluded from tracing. A trace will be excluded if its root span contains a http.url tag and if this tag matches any of the provided regular expression using the standard python regexp match semantic (https://docs.python.org/2/library/re.html#re.match).

Parameters: regexps (list) – a list of regular expressions (or a single string) defining the urls that should be filtered out.

Examples: To filter out http calls to domain api.example.com:

FilterRequestsOnUrl(r'http://api\\.example\\.com')

To filter out http calls to all first level subdomains from example.com:

FilterRequestOnUrl(r'http://.*+\\.example\\.com')

To filter out calls to both http://test.example.com and http://example.com/healthcheck:

FilterRequestOnUrl([r'http://test\\.example\\.com', r'http://example\\.com/healthcheck'])

process_trace(trace: List[Span]) → Optional[List[Span]]¶: When the filter is registered in the tracer, process_trace is called by on each trace before it is sent to the agent, the returned value will be fed to the next filter in the list. If process_trace returns None, the whole trace is discarded.

Writing a custom filter

Create a filter by implementing a class with a process_trace method and providing it to the filters parameter of ddtrace.Tracer.configure(). process_trace should either return a trace to be fed to the next step of the pipeline or None if the trace should be discarded:

from ddtrace import Span, tracer
from ddtrace.filters import TraceFilter

class FilterExample(TraceFilter):
    def process_trace(self, trace):
        # type: (List[Span]) -> Optional[List[Span]]
        ...

# And then configure it with
tracer.configure(settings={'FILTERS': [FilterExample()]})

(see filters.py for other example implementations)

Logs Injection¶

Datadog APM traces can be integrated with the logs product by:

1. Having ddtrace patch the logging module. This will add trace attributes to the log record.

2. Updating the log formatter used by the application. In order to inject tracing information into a log the formatter must be updated to include the tracing attributes from the log record. ddtrace-run will do this automatically for you by specifying a format. For more detail or instructions for how to do this manually see the manual section below.

With these in place the trace information will be injected into a log entry which can be used to correlate the log and trace in Datadog.

ddtrace-run¶

When using ddtrace-run, enable patching by setting the environment variable DD_LOGS_INJECTION=true. The logger by default will have a format that includes trace information:

import logging
from ddtrace import tracer

log = logging.getLogger()
log.level = logging.INFO


@tracer.wrap()
def hello():
    log.info('Hello, World!')

hello()

Manual Instrumentation¶

If you prefer to instrument manually, patch the logging library then update the log formatter as in the following example

Make sure that your log format exactly matches the following:

from ddtrace import patch_all; patch_all(logging=True)
import logging
from ddtrace import tracer

FORMAT = ('%(asctime)s %(levelname)s [%(name)s] [%(filename)s:%(lineno)d] '
          '[dd.service=%(dd.service)s dd.env=%(dd.env)s '
          'dd.version=%(dd.version)s '
          'dd.trace_id=%(dd.trace_id)s dd.span_id=%(dd.span_id)s]'
          '- %(message)s')
logging.basicConfig(format=FORMAT)
log = logging.getLogger()
log.level = logging.INFO


@tracer.wrap()
def hello():
    log.info('Hello, World!')

hello()

HTTP tagging¶

Query String Tracing¶

It is possible to store the query string of the URL — the part after the ? in your URL — in the url.query.string tag.

Configuration can be provided both at the global level and at the integration level.

Examples:

from ddtrace import config

# Global config
config.http.trace_query_string = True

# Integration level config, e.g. 'falcon'
config.falcon.http.trace_query_string = True

Headers tracing¶

For a selected set of integrations, it is possible to store http headers from both requests and responses in tags.

Configuration can be provided both at the global level and at the integration level.

Examples:

from ddtrace import config

# Global config
config.trace_headers([
    'user-agent',
    'transfer-encoding',
])

# Integration level config, e.g. 'falcon'
config.falcon.http.trace_headers([
    'user-agent',
    'some-other-header',
])

The following rules apply:

headers configuration is based on a whitelist. If a header does not appear in the whitelist, it won’t be traced.
headers configuration is case-insensitive.
if you configure a specific integration, e.g. ‘requests’, then such configuration overrides the default global configuration, only for the specific integration.
if you do not configure a specific integration, then the default global configuration applies, if any.
if no configuration is provided (neither global nor integration-specific), then headers are not traced.

Once you configure your application for tracing, you will have the headers attached to the trace as tags, with a structure like in the following example:

http {
  method  GET
  request {
    headers {
      user_agent  my-app/0.0.1
    }
  }
  response {
    headers {
      transfer_encoding  chunked
    }
  }
  status_code  200
  url  https://api.github.com/events
}

Custom Error Codes¶

It is possible to have a custom mapping of which HTTP status codes are considered errors. By default, 500-599 status codes are considered errors. Configuration is provided both at the global level.

Examples:

from ddtrace import config

config.http_server.error_statuses = '500-599'

Certain status codes can be excluded by providing a list of ranges. Valid options:

400-400
400-403,405-499
400,401,403

OpenTracing¶

The Datadog opentracer can be configured via the config dictionary parameter to the tracer which accepts the following described fields. See below for usage.

Configuration Key	Description	Default Value
enabled	enable or disable the tracer	True
debug	enable debug logging	False
agent_hostname	hostname of the Datadog agent to use	localhost
agent_https	use https to connect to the agent	False
agent_port	port the Datadog agent is listening on	8126
global_tags	tags that will be applied to each span	{}
uds_path	unix socket of agent to connect to	None
settings	see Advanced Usage	{}

Usage¶

Manual tracing

To explicitly trace:

import time
import opentracing
from ddtrace.opentracer import Tracer, set_global_tracer

def init_tracer(service_name):
    config = {
      'agent_hostname': 'localhost',
      'agent_port': 8126,
    }
    tracer = Tracer(service_name, config=config)
    set_global_tracer(tracer)
    return tracer

def my_operation():
  span = opentracing.tracer.start_span('my_operation_name')
  span.set_tag('my_interesting_tag', 'my_interesting_value')
  time.sleep(0.05)
  span.finish()

init_tracer('my_service_name')
my_operation()

Context Manager Tracing

To trace a function using the span context manager:

import time
import opentracing
from ddtrace.opentracer import Tracer, set_global_tracer

def init_tracer(service_name):
    config = {
      'agent_hostname': 'localhost',
      'agent_port': 8126,
    }
    tracer = Tracer(service_name, config=config)
    set_global_tracer(tracer)
    return tracer

def my_operation():
  with opentracing.tracer.start_span('my_operation_name') as span:
    span.set_tag('my_interesting_tag', 'my_interesting_value')
    time.sleep(0.05)

init_tracer('my_service_name')
my_operation()

See our tracing trace-examples repository for concrete, runnable examples of the Datadog opentracer.

See also the Python OpenTracing repository for usage of the tracer.

Alongside Datadog tracer

The Datadog OpenTracing tracer can be used alongside the Datadog tracer. This provides the advantage of providing tracing information collected by ddtrace in addition to OpenTracing. The simplest way to do this is to use the ddtrace-run command to invoke your OpenTraced application.

Examples¶

Celery

Distributed Tracing across celery tasks with OpenTracing.

Install Celery OpenTracing:
```
pip install Celery-OpenTracing
```

Replace your Celery app with the version that comes with Celery-OpenTracing:

from celery_opentracing import CeleryTracing
from ddtrace.opentracer import set_global_tracer, Tracer

ddtracer = Tracer()
set_global_tracer(ddtracer)

app = CeleryTracing(app, tracer=ddtracer)

Opentracer API¶

class ddtrace.opentracer.Tracer(service_name=None, config=None, scope_manager=None, dd_tracer=None)¶

A wrapper providing an OpenTracing API for the Datadog tracer.

__init__(service_name=None, config=None, scope_manager=None, dd_tracer=None)¶

Initialize a new Datadog opentracer.

Parameters

service_name – (optional) the name of the service that this tracer will be used with. Note if not provided, a service name will try to be determined based off of sys.argv. If this fails a ddtrace.settings.ConfigException will be raised.
config – (optional) a configuration object to specify additional options. See the documentation for further information.
scope_manager – (optional) the scope manager for this tracer to use. The available managers are listed in the Python OpenTracing repo here: https://github.com/opentracing/opentracing-python#scope-managers. If None is provided, defaults to opentracing.scope_managers.ThreadLocalScopeManager.
dd_tracer – (optional) the Datadog tracer for this tracer to use. This should only be passed if a custom Datadog tracer is being used. Defaults to the global ddtrace.tracer tracer.

property scope_manager¶: Returns the scope manager being used by this tracer.

start_active_span(operation_name, child_of=None, references=None, tags=None, start_time=None, ignore_active_span=False, finish_on_close=True)¶

Returns a newly started and activated Scope. The returned Scope supports with-statement contexts. For example:

with tracer.start_active_span('...') as scope:
    scope.span.set_tag('http.method', 'GET')
    do_some_work()
# Span.finish() is called as part of Scope deactivation through
# the with statement.

It’s also possible to not finish the Span when the Scope context expires:

with tracer.start_active_span('...',
                              finish_on_close=False) as scope:
    scope.span.set_tag('http.method', 'GET')
    do_some_work()
# Span.finish() is not called as part of Scope deactivation as
# `finish_on_close` is `False`.

Parameters

operation_name – name of the operation represented by the new span from the perspective of the current service.
child_of – (optional) a Span or SpanContext instance representing the parent in a REFERENCE_CHILD_OF Reference. If specified, the references parameter must be omitted.
references – (optional) a list of Reference objects that identify one or more parent SpanContexts. (See the Reference documentation for detail).
tags – an optional dictionary of Span Tags. The caller gives up ownership of that dictionary, because the Tracer may use it as-is to avoid extra data copying.
start_time – an explicit Span start time as a unix timestamp per time.time().
ignore_active_span – (optional) an explicit flag that ignores the current active Scope and creates a root Span.
finish_on_close – whether span should automatically be finished when Scope.close() is called.

Returns

a Scope, already registered via the ScopeManager.

start_span(operation_name=None, child_of=None, references=None, tags=None, start_time=None, ignore_active_span=False)¶

Starts and returns a new Span representing a unit of work.

Starting a root Span (a Span with no causal references):

tracer.start_span('...')

Starting a child Span (see also start_child_span()):

tracer.start_span(
    '...',
    child_of=parent_span)

Starting a child Span in a more verbose way:

tracer.start_span(
    '...',
    references=[opentracing.child_of(parent_span)])

Note: the precedence when defining a relationship is the following, from highest to lowest: 1. child_of 2. references 3. scope_manager.active (unless ignore_active_span is True) 4. None

Currently Datadog only supports child_of references.

Parameters

operation_name – name of the operation represented by the new span from the perspective of the current service.
child_of – (optional) a Span or SpanContext instance representing the parent in a REFERENCE_CHILD_OF Reference. If specified, the references parameter must be omitted.
references – (optional) a list of Reference objects that identify one or more parent SpanContexts. (See the Reference documentation for detail)
tags – an optional dictionary of Span Tags. The caller gives up ownership of that dictionary, because the Tracer may use it as-is to avoid extra data copying.
start_time – an explicit Span start time as a unix timestamp per time.time()
ignore_active_span – an explicit flag that ignores the current active Scope and creates a root Span.

Returns

an already-started Span instance.

property active_span¶

Retrieves the active span from the opentracing scope manager

Falls back to using the datadog active span if one is not found. This allows opentracing users to use datadog instrumentation.

inject(span_context, format, carrier)¶

Injects a span context into a carrier.

Parameters

span_context – span context to inject.
format – format to encode the span context with.
carrier – the carrier of the encoded span context.

extract(format, carrier)¶

Extracts a span context from a carrier.

Parameters

format – format that the carrier is encoded with.
carrier – the carrier to extract from.

`ddtrace-run`¶

ddtrace-run will trace supported web frameworks and database modules without the need for changing your code:

$ ddtrace-run -h

Execute the given Python program, after configuring it
to emit Datadog traces.

Append command line arguments to your program as usual.

Usage: ddtrace-run <my_program>

The environment variables for ddtrace-run used to configure the tracer are detailed in Configuration.

ddtrace-run respects a variety of common entrypoints for web applications:

ddtrace-run python my_app.py
ddtrace-run python manage.py runserver
ddtrace-run gunicorn myapp.wsgi:application

Pass along command-line arguments as your program would normally expect them:

$ ddtrace-run gunicorn myapp.wsgi:application --max-requests 1000 --statsd-host localhost:8125

If you’re running in a Kubernetes cluster and still don’t see your traces, make sure your application has a route to the tracing Agent. An easy way to test this is with a:

$ pip install ipython
$ DD_TRACE_DEBUG=true ddtrace-run ipython

Because iPython uses SQLite, it will be automatically instrumented and your traces should be sent off. If an error occurs, a message will be displayed in the console, and changes can be made as needed.

uWSGI¶

The tracer and profiler support uWSGI when configured with the following:

Threads must be enabled with enable-threads or with threads if running uWSGI in multithreaded mode.
If manual instrumentation and configuration is used, lazy-apps must be used.

To enable tracing with automatic instrumentation and configuration with environment variables, use import option with the setting ddtrace.bootstrap.customize. For example, add the following to the uWSGI configuration file:

import=ddtrace.bootstrap.sitecustomize

Note: Automatic instrumentation and configuration using ddtrace-run is not supported with uWSGI.

To enable tracing with manual instrumentation and configuration, configure uWSGI with the lazy-apps option and use patch_all() and agent configuration to a WSGI app:

from ddtrace import patch_all
from ddtrace import tracer


patch_all()
tracer.configure()

def application(env, start_response):
    with tracer.trace("uwsgi-app"):
        start_response('200 OK', [('Content-Type','text/html')])
        return [b"Hello World"]

Gunicorn¶

ddtrace supports Gunicorn.

However, if you are using the gevent worker class, you have to make sure gevent monkey patching is done before loading the ddtrace library.

There are different options to make that happen:

If you rely on ddtrace-run, you must set DD_GEVENT_PATCH_ALL=1 in your environment to have gevent patched first-thing.
Replace ddtrace-run by using import ddtrace.bootstrap.sitecustomize as the first import of your application.
Use a post_worker_init hook to import ddtrace.bootstrap.sitecustomize.

API¶

`Tracer`¶

class ddtrace.Tracer(url: Optional[str] = None, dogstatsd_url: Optional[str] = None)¶

Tracer is used to create, sample and submit spans that measure the execution time of sections of code.

If you’re running an application that will serve a single trace per thread, you can use the global tracer instance:

from ddtrace import tracer
trace = tracer.trace('app.request', 'web-server').finish()

__init__(url: Optional[str] = None, dogstatsd_url: Optional[str] = None) → None¶

Create a new Tracer instance. A global tracer is already initialized for common usage, so there is no need to initialize your own Tracer.

Parameters

url – The Datadog agent URL.
dogstatsd_url – The DogStatsD URL.

on_start_span(func: Callable) → Callable¶

Register a function to execute when a span start.

Can be used as a decorator.

Parameters: func – The function to call when starting a span. The started span will be passed as argument.

deregister_on_start_span(func: Callable) → Callable¶

Unregister a function registered to execute when a span starts.

Can be used as a decorator.

Parameters: func – The function to stop calling when starting a span.

global_excepthook(tp, value, traceback)¶: The global tracer except hook.

get_call_context(*args, **kwargs) → ddtrace.context.Context¶

Return the current active Context for this traced execution. This method is automatically called in the tracer.trace(), but it can be used in the application code during manual instrumentation like:

from ddtrace import tracer

async def web_handler(request):
    context = tracer.get_call_context()
    # use the context if needed
    # ...

This method makes use of a ContextProvider that is automatically set during the tracer initialization, or while using a library instrumentation.

current_trace_context(*args, **kwargs)¶

Return the context for the current trace.

If there is no active trace then None is returned.

configure(enabled: Optional[bool] = None, hostname: Optional[str] = None, port: Optional[int] = None, uds_path: Optional[str] = None, https: Optional[bool] = None, sampler: Optional[ddtrace.sampler.BaseSampler] = None, context_provider: Optional[ddtrace.provider.DefaultContextProvider] = None, wrap_executor: Optional[Callable] = None, priority_sampling: Optional[bool] = None, settings: Optional[Dict[str, Any]] = None, dogstatsd_url: Optional[str] = None, writer: Optional[ddtrace.internal.writer.TraceWriter] = None, partial_flush_enabled: Optional[bool] = None, partial_flush_min_spans: Optional[int] = None) → None¶

Configure an existing Tracer the easy way. Allow to configure or reconfigure a Tracer instance.

Parameters

enabled (bool) – If True, finished traces will be submitted to the API. Otherwise they’ll be dropped.
hostname (str) – Hostname running the Trace Agent
port (int) – Port of the Trace Agent
uds_path (str) – The Unix Domain Socket path of the agent.
https (bool) – Whether to use HTTPS or HTTP.
sampler (object) – A custom Sampler instance, locally deciding to totally drop the trace or not.
context_provider (object) – The ContextProvider that will be used to retrieve automatically the current call context. This is an advanced option that usually doesn’t need to be changed from the default value
wrap_executor (object) – callable that is used when a function is decorated with Tracer.wrap(). This is an advanced option that usually doesn’t need to be changed from the default value
priority_sampling – enable priority sampling, this is required for complete distributed tracing support. Enabled by default.
dogstatsd_url (str) – URL for UDP or Unix socket connection to DogStatsD

start_span(name: str, child_of: Optional[Union[ddtrace.span.Span, ddtrace.context.Context]] = None, service: Optional[str] = None, resource: Optional[str] = None, span_type: Optional[str] = None, activate: bool = False) → ddtrace.span.Span¶

Return a span that represents an operation called name.

Note that the trace() method will almost always be preferred over this method as it provides automatic span parenting. This method should only be used if manual parenting is desired.

Parameters

name (str) – the name of the operation being traced.
child_of (object) – a Span or a Context instance representing the parent for this span.
service (str) – the name of the service being traced.
resource (str) – an optional name of the resource being tracked.
span_type (str) – an optional operation type.
activate – activate the span once it is created.

To start a new root span:

span = tracer.start_span("web.request")

To create a child for a root span:

root_span = tracer.start_span("web.request")
span = tracer.start_span("web.decoder", child_of=root_span)

Spans from start_span are not activated by default:

with tracer.start_span("parent") as parent:
    assert tracer.current_span() is None
    with tracer.start_span("child", child_of=parent):
        assert tracer.current_span() is None

new_parent = tracer.start_span("new_parent", activate=True)
assert tracer.current_span() is new_parent

Note: be sure to finish all spans to avoid memory leaks and incorrect parenting of spans.

trace(name: str, service: Optional[str] = None, resource: Optional[str] = None, span_type: Optional[str] = None) → ddtrace.span.Span¶

Activate and return a new span that inherits from the current active span.

Parameters

name (str) – the name of the operation being traced
service (str) – the name of the service being traced. If not set, it will inherit the service from its parent.
resource (str) – an optional name of the resource being tracked.
span_type (str) – an optional operation type.

The returned span must be finish’d or it will remain in memory indefinitely:

>>> span = tracer.trace("web.request")
    try:
        # do something
    finally:
        span.finish()

>>> with tracer.trace("web.request") as span:
        # do something

Example of the automatic parenting:

parent = tracer.trace("parent")     # has no parent span
assert tracer.current_span() is parent

child  = tracer.trace("child")
assert child.parent_id == parent.span_id
assert tracer.current_span() is child
child.finish()

# parent is now the active span again
assert tracer.current_span() is parent
parent.finish()

assert tracer.current_span() is None

parent2 = tracer.trace("parent2")
assert parent2.parent_id is None
parent2.finish()

current_root_span() → Optional[ddtrace.span.Span]¶

Returns the root span of the current execution.

This is useful for attaching information related to the trace as a whole without needing to add to child spans.

For example:

# get the root span
root_span = tracer.current_root_span()
# set the host just once on the root span
if root_span:
    root_span.set_tag('host', '127.0.0.1')

current_span() → Optional[ddtrace.span.Span]¶

Return the active span in the current execution context.

Note that there may be an active span represented by a context object (like from a distributed trace) which will not be returned by this method.

write(spans: Optional[List[ddtrace.span.Span]]) → None¶: Send the trace to the writer to enqueue the spans list in the agent sending queue.

set_service_info(*args, **kwargs)¶: Set the information about the given service.

wrap(name: Optional[str] = None, service: Optional[str] = None, resource: Optional[str] = None, span_type: Optional[str] = None) → Callable[[ddtrace.tracer.AnyCallable], ddtrace.tracer.AnyCallable]¶

A decorator used to trace an entire function. If the traced function is a coroutine, it traces the coroutine execution when is awaited. If a wrap_executor callable has been provided in the Tracer.configure() method, it will be called instead of the default one when the function decorator is invoked.

Parameters

name (str) – the name of the operation being traced. If not set, defaults to the fully qualified function name.
service (str) – the name of the service being traced. If not set, it will inherit the service from it’s parent.
resource (str) – an optional name of the resource being tracked.
span_type (str) – an optional operation type.

>>> @tracer.wrap('my.wrapped.function', service='my.service')
    def run():
        return 'run'

>>> # name will default to 'execute' if unset
    @tracer.wrap()
    def execute():
        return 'executed'

>>> # or use it in asyncio coroutines
    @tracer.wrap()
    async def coroutine():
        return 'executed'

>>> @tracer.wrap()
    @asyncio.coroutine
    def coroutine():
        return 'executed'

You can access the current span using tracer.current_span() to set tags:

>>> @tracer.wrap()
    def execute():
        span = tracer.current_span()
        span.set_tag('a', 'b')

set_tags(tags: Dict[str, str]) → None¶

Set some tags at the tracer level. This will append those tags to each span created by the tracer.

Parameters: tags (dict) – dict of tags to set at tracer level

shutdown(timeout: Optional[float] = None) → None¶

Shutdown the tracer.

This will stop the background writer/worker and flush any finished traces in the buffer. The tracer cannot be used for tracing after this method has been called. A new tracer instance is required to continue tracing.

Parameters: timeout (int | float | None) – How long in seconds to wait for the background worker to flush traces before exiting or None to block until flushing has successfully completed (default: None)

`Span`¶

class ddtrace.Span(tracer: Optional[Tracer], name: str, service: Optional[str] = None, resource: Optional[str] = None, span_type: Optional[str] = None, trace_id: Optional[int] = None, span_id: Optional[int] = None, parent_id: Optional[int] = None, start: Optional[int] = None, context: Optional[Context] = None, on_finish: Optional[List[Callable[[Span], None]]] = None)¶

__init__(tracer: Optional[Tracer], name: str, service: Optional[str] = None, resource: Optional[str] = None, span_type: Optional[str] = None, trace_id: Optional[int] = None, span_id: Optional[int] = None, parent_id: Optional[int] = None, start: Optional[int] = None, context: Optional[Context] = None, on_finish: Optional[List[Callable[[Span], None]]] = None) → None¶

Create a new span. Call finish once the traced operation is over.

Parameters

tracer (ddtrace.Tracer) – the tracer that will submit this span when finished.
name (str) – the name of the traced operation.
service (str) – the service name
resource (str) – the resource name
span_type (str) – the span type
trace_id (int) – the id of this trace’s root span.
parent_id (int) – the id of this span’s direct parent span.
span_id (int) – the id of this span.
start (int) – the start time of request as a unix epoch in seconds
context (object) – the Context of the span.
on_finish – list of functions called when the span finishes.

property start¶: The start timestamp in Unix epoch seconds.

property duration¶: The span duration in seconds.

finish(finish_time: Optional[float] = None) → None¶

Mark the end time of the span and submit it to the tracer. If the span has already been finished don’t do anything.

Parameters: finish_time – The end time of the span, in seconds. Defaults to now.

set_tag(key: Union[str, bytes], value: Optional[Any] = None) → None¶

Set a tag key/value pair on the span.

Keys must be strings, values must be stringify-able.

Parameters

key (str) – Key to use for the tag
value (stringify-able value) – Value to assign for the tag

get_tag(key: Union[str, bytes]) → Optional[str]¶: Return the given tag or None if it doesn’t exist.

set_tags(tags: Dict[Union[str, bytes], str]) → None¶: Set a dictionary of tags on the given span. Keys and values must be strings (or stringable)

set_traceback(limit: int = 20) → None¶: If the current stack has an exception, tag the span with the relevant error info. If not, set the span to the current python stack.

set_exc_info(exc_type: Any, exc_val: Any, exc_tb: Any) → None¶: Tag the span with an error tuple as from sys.exc_info().

pprint() → str¶: Return a human readable version of the span.

property context¶: Return the trace context for this span.

`Context`¶

class ddtrace.context.Context(trace_id: Optional[int] = None, span_id: Optional[int] = None, dd_origin: Optional[str] = None, sampling_priority: Optional[Union[int, float]] = None, lock: threading.RLock = NOTHING, meta=NOTHING, metrics=NOTHING)¶

Represents the state required to propagate a trace across execution boundaries.

property sampling_priority¶: Return the context sampling priority for the trace.

__init__(trace_id: Optional[int] = None, span_id: Optional[int] = None, dd_origin: Optional[str] = None, sampling_priority: Optional[Union[int, float]] = None, lock: threading.RLock = NOTHING, meta=NOTHING, metrics=NOTHING) → None¶: Method generated by attrs for class Context.

property dd_origin¶: Get the origin of the trace.

clone() → ddtrace.context.Context¶: Partially clones the current context. It copies everything EXCEPT the registered and finished spans.

`Pin`¶

class ddtrace.Pin(service: Optional[str] = None, app: Optional[str] = None, app_type=None, tags: Optional[Dict[str, str]] = None, tracer: Optional[Tracer] = None, _config: Optional[Dict[str, Any]] = None)¶

Pin (a.k.a Patch INfo) is a small class which is used to set tracing metadata on a particular traced connection. This is useful if you wanted to, say, trace two different database clusters.

>>> conn = sqlite.connect('/tmp/user.db')
>>> # Override a pin for a specific connection
>>> pin = Pin.override(conn, service='user-db')
>>> conn = sqlite.connect('/tmp/image.db')

property service¶: Backward compatibility: accessing to pin.service returns the underlying configuration value.

static get_from(obj: Any) → ddtrace.pin.Pin¶

Return the pin associated with the given object. If a pin is attached to obj but the instance is not the owner of the pin, a new pin is cloned and attached. This ensures that a pin inherited from a class is a copy for the new instance, avoiding that a specific instance overrides other pins values.

>>> pin = Pin.get_from(conn)

Parameters: obj (object) – The object to look for a ddtrace.pin.Pin on
Return type: ddtrace.pin.Pin, None
Returns: ddtrace.pin.Pin associated with the object, or None if none was found

classmethod override(obj: Any, service: Optional[str] = None, app: Optional[str] = None, app_type=None, tags: Optional[Dict[str, str]] = None, tracer: Optional[Tracer] = None) → None¶

Override an object with the given attributes.

That’s the recommended way to customize an already instrumented client, without losing existing attributes.

>>> conn = sqlite.connect('/tmp/user.db')
>>> # Override a pin for a specific connection
>>> Pin.override(conn, service='user-db')

enabled() → bool¶: Return true if this pin’s tracer is enabled.

onto(obj: Any, send: bool = True) → None¶: Patch this pin onto the given object. If send is true, it will also queue the metadata to be sent to the server.

clone(service: Optional[str] = None, app: Optional[str] = None, app_type=None, tags: Optional[Dict[str, str]] = None, tracer: Optional[Tracer] = None) → Pin¶: Return a clone of the pin with the given attributes replaced.

`patch_all`¶

ddtrace.monkey.patch_all(**patch_modules: Dict[str, bool]) → None¶

Automatically patches all available modules.

In addition to patch_modules, an override can be specified via an environment variable, DD_TRACE_<module>_ENABLED for each module.

patch_modules have the highest precedence for overriding.

Parameters

patch_modules (dict) –

Override whether particular modules are patched or not.

>>> patch_all(redis=False, cassandra=False)

`patch`¶

ddtrace.monkey.patch(raise_errors: bool = True, **patch_modules: Dict[str, bool]) → None¶

Patch only a set of given modules.

Parameters

raise_errors (bool) – Raise error if one patch fail.
patch_modules (dict) –
List of modules to patch.
```
>>> patch(psycopg=True, elasticsearch=True)
```

Advanced Usage¶

Agent Configuration¶

Context¶

Tracing Context Management¶

Tracing Across Threads¶

Tracing Across Processes¶

fork¶

Tracing Across Asyncio Tasks¶

Manual Management¶

Context Providers¶

Distributed Tracing¶

Web Frameworks¶

HTTP Client¶

Custom¶

Resolving deprecation warnings¶

Trace Filtering¶

Logs Injection¶

ddtrace-run¶

Manual Instrumentation¶

HTTP tagging¶

Query String Tracing¶

Headers tracing¶

Custom Error Codes¶

OpenTracing¶

Usage¶

Examples¶

Opentracer API¶

ddtrace-run¶

uWSGI¶

Gunicorn¶

API¶

Tracer¶

Span¶

Context¶

Pin¶

patch_all¶

patch¶

`ddtrace-run`¶

`Tracer`¶

`Span`¶

`Context`¶

`Pin`¶

`patch_all`¶

`patch`¶