Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Automatic exporter/provider setup for opentelemetry-instrument command. #1036

Merged
merged 1 commit into from
Nov 20, 2020
Merged
Show file tree
Hide file tree
Changes from all commits
Commits
File filter

Filter by extension

Filter by extension

Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
4 changes: 4 additions & 0 deletions exporter/opentelemetry-exporter-jaeger/setup.cfg
Original file line number Diff line number Diff line change
Expand Up @@ -48,3 +48,7 @@ where = src

[options.extras_require]
test =

[options.entry_points]
opentelemetry_exporter =
jaeger = opentelemetry.exporter.jaeger:JaegerSpanExporter
4 changes: 4 additions & 0 deletions exporter/opentelemetry-exporter-opencensus/setup.cfg
Original file line number Diff line number Diff line change
Expand Up @@ -50,3 +50,7 @@ where = src

[options.extras_require]
test =

[options.entry_points]
opentelemetry_exporter =
opencensus = opentelemetry.exporter.opencensus.trace_exporter:OpenCensusSpanExporter
5 changes: 5 additions & 0 deletions exporter/opentelemetry-exporter-otlp/setup.cfg
Original file line number Diff line number Diff line change
Expand Up @@ -52,3 +52,8 @@ test =

[options.packages.find]
where = src

[options.entry_points]
opentelemetry_exporter =
otlp_span = opentelemetry.exporter.otlp.trace_exporter:OTLPSpanExporter
otlp_metric = opentelemetry.exporter.otlp.metrics_exporter:OTLPMetricsExporter
4 changes: 4 additions & 0 deletions exporter/opentelemetry-exporter-prometheus/setup.cfg
Original file line number Diff line number Diff line change
Expand Up @@ -48,3 +48,7 @@ where = src

[options.extras_require]
test =

[options.entry_points]
opentelemetry_exporter =
prometheus = opentelemetry.exporter.prometheus:PrometheusMetricsExporter
4 changes: 4 additions & 0 deletions exporter/opentelemetry-exporter-zipkin/setup.cfg
Original file line number Diff line number Diff line change
Expand Up @@ -48,3 +48,7 @@ where = src

[options.extras_require]
test =

[options.entry_points]
opentelemetry_exporter =
zipkin = opentelemetry.exporter.zipkin:ZipkinSpanExporter
4 changes: 3 additions & 1 deletion opentelemetry-instrumentation/CHANGELOG.md
Original file line number Diff line number Diff line change
Expand Up @@ -2,11 +2,13 @@

## Unreleased

- Added support for `OTEL_EXPORTER` to the `opentelemetry-instrument` command ([#1036](https://github.com/open-telemetry/opentelemetry-python/pull/1036))

## Version 0.14b0

Released 2020-10-13

- Fixed boostrap command to correctly install opentelemetry-instrumentation-falcon instead of opentelemetry-instrumentation-flask
- Fixed bootstrap command to correctly install opentelemetry-instrumentation-falcon instead of opentelemetry-instrumentation-flask. ([#1138](https://github.com/open-telemetry/opentelemetry-python/pull/1138))

## Version 0.13b0

Expand Down
68 changes: 60 additions & 8 deletions opentelemetry-instrumentation/README.rst
Original file line number Diff line number Diff line change
Expand Up @@ -16,30 +16,82 @@ Installation

This package provides a couple of commands that help automatically instruments a program:


opentelemetry-bootstrap
-----------------------

::

opentelemetry-bootstrap --action=install|requirements

This commands inspects the active Python site-packages and figures out which
instrumentation packages the user might want to install. By default it prints out
a list of the suggested instrumentation packages which can be added to a requirements.txt
file. It also supports installing the suggested packages when run with :code:`--action=install`
flag.


opentelemetry-instrument
------------------------

::

opentelemetry-instrument python program.py

The instrument command will try to automatically detect packages used by your python program
and when possible, apply automatic tracing instrumentation on them. This means your program
will get automatic distributed tracing for free without having to make any code changes
at all. This will also configure a global tracer and tracing exporter without you having to
make any code changes. By default, the instrument command will use the OTLP exporter but
this can be overriden when needed.

The command supports the following configuration options as CLI arguments and environment vars:


* ``--exporter`` or ``OTEL_EXPORTER``

Used to specify which trace exporter to use. Can be set to one or more
of the well-known exporter names (see below).

- Defaults to `otlp`.
- Can be set to `none` to disable automatic tracer initialization.

You can pass multiple values to configure multiple exporters e.g, ``zipkin,prometheus``

Well known trace exporter names:

- jaeger
- opencensus
- otlp
- otlp_span
- otlp_metric
- zipkin

``otlp`` is an alias for ``otlp_span,otlp_metric``.

* ``--service-name`` or ``OTEL_SERVICE_NAME``

When present the value is passed on to the relevant exporter initializer as ``service_name`` argument.

The code in ``program.py`` needs to use one of the packages for which there is
an OpenTelemetry integration. For a list of the available integrations please
check `here <https://opentelemetry-python.readthedocs.io/en/stable/index.html#integrations>`_

Examples
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^

opentelemetry-bootstrap
-----------------------
::

opentelemetry-instrument -e otlp flask run --port=3000

The above command will pass ``-e otlp`` to the instrument command and ``--port=3000`` to ``flask run``.

::

opentelemetry-bootstrap --action=install|requirements
opentelemetry-instrument -e zipkin,otlp celery -A tasks worker --loglevel=info
Copy link
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

can you explain the reason behind removing the info about opentelemetry-bootstrap?

Copy link
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

+1, this info is probably important to keep here.

Copy link
Contributor Author

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

This is unintentional. Probably caused by a bad rebase. Fixing.


This commands inspects the active Python site-packages and figures out which
instrumentation packages the user might want to install. By default it prints out
a list of the suggested instrumentation packages which can be added to a requirements.txt
file. It also supports installing the suggested packages when run with :code:`--action=install`
flag.
The above command will configure global trace provider, attach zipkin and otlp exporters to it and then
start celery with the rest of the arguments.

References
----------
Expand Down
Original file line number Diff line number Diff line change
Expand Up @@ -14,16 +14,67 @@
# See the License for the specific language governing permissions and
# limitations under the License.

import argparse
from logging import getLogger
from os import environ, execl, getcwd
from os.path import abspath, dirname, pathsep
from shutil import which
from sys import argv

logger = getLogger(__file__)


def parse_args():
parser = argparse.ArgumentParser(
description="""
opentelemetry-instrument automatically instruments a Python
program and it's dependencies and then runs the program.
"""
)

parser.add_argument(
"-e",
"--exporter",
required=False,
help="""
Uses the specified exporter to export spans or metrics.
Accepts multiple exporters as comma separated values.

Examples:

-e=otlp
-e=otlp_span,prometheus
-e=jaeger,otlp_metric
""",
)

parser.add_argument(
"-s",
"--service-name",
required=False,
help="""
The service name that should be passed to a trace exporter.
""",
)

parser.add_argument("command", help="Your Python application.")
parser.add_argument(
"command_args",
help="Arguments for your application.",
nargs=argparse.REMAINDER,
)
return parser.parse_args()


def load_config_from_cli_args(args):
if args.exporter:
environ["OTEL_EXPORTER"] = args.exporter
if args.service_name:
environ["OTEL_SERVICE_NAME"] = args.service_name


def run() -> None:
args = parse_args()
load_config_from_cli_args(args)

python_path = environ.get("PYTHONPATH")

Expand All @@ -49,6 +100,5 @@ def run() -> None:

environ["PYTHONPATH"] = pathsep.join(python_path)

executable = which(argv[1])

execl(executable, executable, *argv[2:])
executable = which(args.command)
execl(executable, executable, *args.command_args)
Original file line number Diff line number Diff line change
@@ -0,0 +1,121 @@
# Copyright The OpenTelemetry Authors
#
# Licensed under the Apache License, Version 2.0 (the "License");
# you may not use this file except in compliance with the License.
# You may obtain a copy of the License at
#
# http://www.apache.org/licenses/LICENSE-2.0
#
# Unless required by applicable law or agreed to in writing, software
# distributed under the License is distributed on an "AS IS" BASIS,
# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
# See the License for the specific language governing permissions and
# limitations under the License.

from logging import getLogger
from typing import Sequence, Tuple

from pkg_resources import iter_entry_points

from opentelemetry import trace
from opentelemetry.configuration import Configuration
from opentelemetry.sdk.metrics.export import MetricsExporter
from opentelemetry.sdk.resources import Resource
from opentelemetry.sdk.trace import TracerProvider
from opentelemetry.sdk.trace.export import (
BatchExportSpanProcessor,
SpanExporter,
)

logger = getLogger(__file__)

EXPORTER_OTLP = "otlp"
EXPORTER_OTLP_SPAN = "otlp_span"
EXPORTER_OTLP_METRIC = "otlp_metric"
_DEFAULT_EXPORTER = EXPORTER_OTLP


def get_service_name() -> str:
return Configuration().SERVICE_NAME or ""


def get_exporter_names() -> Sequence[str]:
exporter = Configuration().EXPORTER or _DEFAULT_EXPORTER
if exporter.lower().strip() == "none":
return []

names = []
for exp in exporter.split(","):
name = exp.strip()
if name == EXPORTER_OTLP:
names.append(EXPORTER_OTLP_SPAN)
names.append(EXPORTER_OTLP_METRIC)
else:
names.append(name)
return names


def init_tracing(exporters: Sequence[SpanExporter]):
service_name = get_service_name()
provider = TracerProvider(
resource=Resource.create({"service.name": service_name}),
)
trace.set_tracer_provider(provider)

for exporter_name, exporter_class in exporters.items():
exporter_args = {}
if exporter_name not in [
EXPORTER_OTLP,
EXPORTER_OTLP_SPAN,
]:
exporter_args["service_name"] = service_name

provider.add_span_processor(
BatchExportSpanProcessor(exporter_class(**exporter_args))
)


def init_metrics(exporters: Sequence[MetricsExporter]):
if exporters:
logger.warning("automatic metric initialization is not supported yet.")


def import_exporters(
exporter_names: Sequence[str],
) -> Tuple[Sequence[SpanExporter], Sequence[MetricsExporter]]:
trace_exporters, metric_exporters = {}, {}

exporters = {
ep.name: ep for ep in iter_entry_points("opentelemetry_exporter")
}

for exporter_name in exporter_names:
entry_point = exporters.get(exporter_name, None)
if not entry_point:
raise RuntimeError(
"Requested exporter not found: {0}".format(exporter_name)
)

exporter_impl = entry_point.load()
if issubclass(exporter_impl, SpanExporter):
trace_exporters[exporter_name] = exporter_impl
elif issubclass(exporter_impl, MetricsExporter):
metric_exporters[exporter_name] = exporter_impl
else:
raise RuntimeError(
"{0} is neither a trace exporter nor a metric exporter".format(
exporter_name
)
)
return trace_exporters, metric_exporters


def initialize_components():
ocelotl marked this conversation as resolved.
Show resolved Hide resolved
exporter_names = get_exporter_names()
trace_exporters, metric_exporters = import_exporters(exporter_names)
init_tracing(trace_exporters)

# We don't support automatic initialization for metric yet but have added
# some boilerplate in order to make sure current implementation does not
# lock us out of supporting metrics later without major surgery.
init_metrics(metric_exporters)
Loading