Custom function based sampler for opentelemetry-python
.
$ pip install opentelemetry-hook-sampler
import opentelemetry.trace
from opentelemetry.sdk.resources import SERVICE_NAME
from opentelemetry.sdk.trace import TracerProvider
from opentelemetry_hook_sampler import HookSampler
def sampler_hook():
"""
Sample 1 in 10 traces.
The example is static, but you can get info from any context available
in your application. E.g. http request, celery task, thread locals, etc.
"""
return 10
resource = Resource(attributes={SERVICE_NAME: "foo"})
sampler = HookSampler(sampler_hook)
provider = TracerProvider(resource=resource, sampler=sampler)
opentelemetry.trace.set_tracer_provider(provider)
Honeycomb expects a
SampleRate
attribute to normalize data. It is not OpenTelemetry spec, but
it is supported through opentelemetry_hook_sampler.HoneycombHookSampler
.
Be careful, threadlocals module don't work with async Django.
from threadlocals.threadlocals import get_current_request
def sampler_hook():
request = get_current_request()
# 10% /foo requests
if request.path == '/foo':
return 10
return 1
from django_g import get_current_request
def sampler_hook():
request = get_current_request()
# 10% /foo requests
if request.path == '/foo':
return 10
return 1
import celery
def sampler_hook():
task_name = celery.current_app.current_worker_task.request.task
# 10% foo.bar.baz tasks
if task_name == 'foo.bar.baz':
return 10
return 1