Prometheus metric target_info with SERVICE_NAME contains a dot

[kornerc]

Describe your environment

• Python = 3.11.4
• opentelemetry-api = 1.21.0
• opentelemetry-sdk = 1.21.0
• opentelemetry-exporter-prometheus = 0.42b0
• prometheus-client = 0.19.0
• Prometheus = 2.48.0
• Windows

Steps to reproduce
When using the Prometheus exporter and initializing the Resource with the attribute opentelemetry.sdk.resources.SERVICE_NAME as suggested in the docs, a metric target_info gets exported with the attribute service.name.
Prometheus does not like the dot . in the attribute name and fails scraping the target.

Minimal code to reproduce the issue:

import time

from prometheus_client import start_http_server

from opentelemetry import metrics
from opentelemetry.exporter.prometheus import PrometheusMetricReader
from opentelemetry.sdk.metrics import MeterProvider
from opentelemetry.sdk.resources import SERVICE_NAME, Resource

# Service name is required for most backends
resource = Resource(attributes={
    SERVICE_NAME: "your-service-name"
})

# Start Prometheus client
start_http_server(port=8000, addr="localhost")
# Initialize PrometheusMetricReader which pulls metrics from the SDK
# on-demand to respond to scrape requests
reader = PrometheusMetricReader()
provider = MeterProvider(resource=resource, metric_readers=[reader])
metrics.set_meter_provider(provider)

# Creates a meter from the global meter provider
meter = metrics.get_meter("my.meter.name")

work_counter = meter.create_counter(
    "work.counter", unit="1", description="Counts the amount of work done"
)

while True:
    work_counter.add(1, {"work.type": "Debug"})
    print("doing some work...")
    time.sleep(1)

When browsing to localhost:8000/metrics these are the metrics:

# HELP python_gc_objects_collected_total Objects collected during gc
# TYPE python_gc_objects_collected_total counter
python_gc_objects_collected_total{generation="0"} 356.0
python_gc_objects_collected_total{generation="1"} 289.0
python_gc_objects_collected_total{generation="2"} 0.0
# HELP python_gc_objects_uncollectable_total Uncollectable objects found during GC
# TYPE python_gc_objects_uncollectable_total counter
python_gc_objects_uncollectable_total{generation="0"} 0.0
python_gc_objects_uncollectable_total{generation="1"} 0.0
python_gc_objects_uncollectable_total{generation="2"} 0.0
# HELP python_gc_collections_total Number of times this generation was collected
# TYPE python_gc_collections_total counter
python_gc_collections_total{generation="0"} 55.0
python_gc_collections_total{generation="1"} 4.0
python_gc_collections_total{generation="2"} 0.0
# HELP python_info Python platform information
# TYPE python_info gauge
python_info{implementation="CPython",major="3",minor="11",patchlevel="4",version="3.11.4"} 1.0
# HELP target_info Target metadata
# TYPE target_info gauge
target_info{service.name="your-service-name"} 1.0
# HELP work_counter_1_total Counts the amount of work done
# TYPE work_counter_1_total counter
work_counter_1_total{work_type="Debug"} 9.0

When checking the status of the target in the Prometheus UI, the following error is displayed: expected equal, got "." ("INVALID") while parsing: "target_info{service." because of the dot ..

What is the expected behavior?
The attribute of the metric target_info should not have a dot . in its name.
E.g. service_name.

What is the actual behavior?
The attribute of the metric target_info has a dot . in its name service.name (target_info{service.name="your-service-name"} 1.0).

[kornerc]

I was able to find the cause of the issue, it is this line

opentelemetry-python/exporter/opentelemetry-exporter-prometheus/src/opentelemetry/exporter/prometheus/__init__.py

Line 382 in 3dfe224

info.add_metric(labels=list(attributes.keys()), value=attributes)

for each key of attributes the method _sanitize has to be called

opentelemetry-python/exporter/opentelemetry-exporter-prometheus/src/opentelemetry/exporter/prometheus/__init__.py

Line 364 in 3dfe224

def _sanitize(self, key: str) -> str:

I'll write a pull request to fix this issue.

[Jdsleppy]

Linking to duplicate #3563 to close it when your PR is merged. Thank you for the PR!