feat: add Metrics middleware using CloudWatch EMF

[heitorlessa]

Issue #, if available: #1

Description of changes:

• Create quick script to fetch env vars
• Rename Lambda functions/resources that cannot accept a Log Group parameter
  • Payment
  • Loyalty
  • Booking
• Explicitly create log groups with expiration
• Extend templates to accept log expiration date
• Enable API Gateway access logs
• Create API Gateway IAM Role to ship logs
  • No longer needed
• EMF: Allow creation of multiple metrics at the end of a function
• EMF: Allow creation of a metric with a different dimension
• EMF: Move Metric Manager into a separate file

• Tests
  • Test Single metric
  • Test Single metric only reports one metric
  • Test multiple metrics
  • Test schema invalidation error
  • Test incorrect Metric unit
  • Test incorrect Metric value
  • Test multiple namespaces
  • Test no metrics
  • Test log_metrics decorator
  • Test log_metrics decorator as standalone (call the decorated fn)
  • Test namespace env var
  • Test log_metric to ensure it raises deprecation error
  • Update log_metric tests
  • Test over 100 metrics for two separate EMF objects
  • Test maximum of dimensions
  • Double check if an EMF object can surpass 1 Mb
• Add deprecation warning for log_metric
• Add initial docs
• Update HISTORY
• Update example

Features

• Create up to 100 metrics using a single EMF object (large JSON blob)
• Create an one off metric with a different dimension (e.g. Cold Start metric using function_version while all other metrics use service_name as a dimension)
• Validate against common mistakes that CloudWatch would otherwise fail silently
• No stack, custom resource, data collection needed — Metrics are created async by CloudWatch EMF

UX

Multiple metrics

Used alongside other middlewares that call the decorated function

from lambda_python_powertools.metrics import Metrics, MetricUnit

metrics = Metrics()
metrics.add_namespace(name="ServerlessAirline")
metrics.add_metric(name="ColdStart", unit="Count", value=1)
metrics.add_dimension(name="service", value="booking")

@metrics.log_metrics
@tracer.capture_lambda_handler
def testing():
    metrics.add_metric(name="BookingConfirmation", unit="Count", value=1)
    some_code()
    return True

def some_code():
    metrics.add_metric(name="some_other_metric", unit=MetricUnit.Seconds, value=1)
    ...

Single metric feature

Context manager with parameters using similar DX as multiple metrics

with single_metric(name="ColdStart", unit=MetricUnit.Count, value=1) as metric:
    metric.add_dimension(name="function_context", value="$LATEST")

By submitting this pull request, I confirm that you can use, modify, copy, and redistribute this contribution, under the terms of your choice.

[jfuss]

Did you test that producing the whl/sdist with poetry build still included all the files you need (for example is the .json files included correctly in the package)?

[heitorlessa]

@marcioemiranda What are your thoughts? I know you're using the lib extensively