feat: add Metrics middleware using CloudWatch EMF

python/.flake8

@@ -1,6 +1,6 @@
 [flake8]
 exclude = docs, .eggs, setup.py, example, .aws-sam
-ignore = E203, E266, W503, BLK100, W291
+ignore = E203, E266, W503, BLK100, W291, I004
 max-line-length = 120
 max-complexity = 18
 

python/.gitignore

@@ -293,4 +293,6 @@ $RECYCLE.BIN/
 
 # Misc
 
-test_report
+test_report
+wheelhouse
+docs

python/HISTORY.md

@@ -1,5 +1,18 @@
 # HISTORY 
 
+## April 9th, 2020
+
+**0.6.3**
+
+* Fix `log_metrics` decorator logic not calling the decorated function, and exception handling
+
+## April 8th, 2020
+
+**0.6.1**
+
+* Introduces Metrics middleware to utilise CloudWatch Embedded Metric Format
+* Adds deprecation warning for `log_metrics`
+
 ## February 20th, 2020
 
 **0.5.0**

python/Makefile

@@ -15,7 +15,7 @@ lint: format
 	poetry run flake8
 
 test:
-	poetry run pytest
+	poetry run pytest -vvv
 
 test-html:
 	poetry run pytest --cov-report html
@@ -25,18 +25,30 @@ pr: lint test
 build: pr
 	poetry run build
 
+docs: dev
+	poetry run pdoc --html --output-dir docs ./aws_lambda_powertools --force
+
+docs-dev:
+	poetry run pdoc --http : aws_lambda_powertools
+
 #
 # Use `poetry version <major>/<minor></patch>` for version bump
 #
-release:
+release-prod:
 	poetry config pypi-token.pypi ${PYPI_TOKEN}
-	@$(MAKE) build
-	poetry publish
-	rm -rf dist
+	poetry publish -n
 
 release-test:
 	poetry config repositories.testpypi https://test.pypi.org/legacy
 	poetry config pypi-token.pypi ${PYPI_TEST_TOKEN}
-	@$(MAKE) build
-	poetry publish --repository testpypi
-	rm -rf dist
+	poetry publish --repository testpypi -n
+
+build-linux-wheels:
+	poetry build
+	docker run --env PLAT=manylinux1_x86_64 --rm -it -v ${PWD}:/io -w /io quay.io/pypa/manylinux1_x86_64 /io/build_linux_wheels.sh
+	cp ./wheelhouse/* dist/ && rm -rf wheelhouse
+
+release:
+	$(MAKE) build-linux-wheels
+	$(MAKE) release-test
+	$(MAKE) release-prod

python/README.md

@@ -25,6 +25,13 @@ A suite of utilities for AWS Lambda Functions that makes tracing with AWS X-Ray,
 * Log sampling enables DEBUG log level for a percentage of requests (disabled by default)
     - Enable via `POWERTOOLS_LOGGER_SAMPLE_RATE=0.1`, ranges from 0 to 1, where 0.1 is 10% and 1 is 100%
 
+**Metrics**
+
+* Aggregate up to 100 metrics using a single CloudWatch Embedded Metric Format object (large JSON blob)
+* Context manager to create an one off metric with a different dimension than metrics already aggregated
+* Validate against common metric definitions mistakes (metric unit, values, max dimensions, max metrics, etc)
+* No stack, custom resource, data collection needed — Metrics are created async by CloudWatch EMF
+
 **Environment variables** used across suite of utilities
 
 Environment variable | Description | Default | Utility
@@ -33,6 +40,7 @@ POWERTOOLS_SERVICE_NAME | Sets service name used for tracing namespace, metrics
 POWERTOOLS_TRACE_DISABLED | Disables tracing | "false" | tracing
 POWERTOOLS_LOGGER_LOG_EVENT | Logs incoming event | "false" | logging
 POWERTOOLS_LOGGER_SAMPLE_RATE | Debug log sampling  | 0 | logging
+POWERTOOLS_METRICS_NAMESPACE | Metrics namespace  | None | metrics
 LOG_LEVEL | Sets logging level | "INFO" | logging
 
 ## Usage
@@ -148,41 +156,60 @@ def handler(event, context)
 
 #### Custom Metrics async
 
-> **NOTE**: This will **likely change after Beta** in light of [new Amazon CloudWatch embedded metric format](https://aws.amazon.com/about-aws/whats-new/2019/11/amazon-cloudwatch-launches-embedded-metric-format/), meaning we won't need an additional stack and interface could change.
+> **NOTE** `log_metric` will be removed once it's GA.
+
+This feature makes use of CloudWatch Embedded Metric Format (EMF) and metrics are created asynchronously by CloudWatch service
+
+> Contrary to `log_metric`, you don't need any custom resource or additional CloudFormation stack anymore.
+
+Metrics middleware validates against the minimum necessary for a metric to be published:
 
-This feature requires [Custom Metrics SAR App](https://serverlessrepo.aws.amazon.com/applications/arn:aws:serverlessrepo:us-east-1:374852340823:applications~async-custom-metrics) in order to process canonical metric lines in CloudWatch Logs. 
+* At least of one Metric and Dimension 
+* Maximum of 9 dimensions
+* Only one Namespace
+* [Any Metric unit supported by CloudWatch](https://docs.aws.amazon.com/AmazonCloudWatch/latest/APIReference/API_MetricDatum.html)
 
-If you're starting from scratch, you may want to see a working example, tune to your needs and deploy within your account - [Serverless Airline Log Processing Stack](https://github.com/aws-samples/aws-serverless-airline-booking/blob/develop/src/backend/log-processing/template.yaml)
+**Creating multiple metrics**
+
+`log_metrics` decorator calls the decorated function, so leave that for last decorator or will fail with `SchemaValidationError` if no metrics are recorded.
 
 ```python
-from aws_lambda_powertools.logging import MetricUnit, log_metric
+from aws_lambda_powertools.metrics import Metrics, MetricUnit
 
-def handler(event, context)
-  log_metric(name="SuccessfulPayment", unit=MetricUnit.Count, value=10, namespace="MyApplication")
-
-  # Optional dimensions
-  log_metric(name="SuccessfulPayment", unit=MetricUnit.Count, value=10, namespace="MyApplication", customer_id="123-abc", charge_id="abc-123")
-
-  # Explicit service name
-  log_metric(service="paymentTest", name="SuccessfulPayment", namespace="MyApplication".....)
-  ...
+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 lambda_handler(evt, ctx):
+    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)
+    ...
 ```
 
-**Exerpt output in CloudWatch Logs**
+CloudWatch EMF uses the same dimensions across all metrics. If you have metrics that should have different dimensions, use `single_metric` to create a single metric with any dimension you want. Generally, this would be an edge case since you [pay for unique metric](https://aws.amazon.com/cloudwatch/pricing/) 
 
-```
-MONITORING|10|Count|SuccessfulPayment|MyApplication|service="payment
-MONITORING|10|Count|SuccessfulPayment|MyApplication|customer_id="123-abc",charge_id="abc-123",service="payment
-MONITORING|10|Count|SuccessfulPayment|MyApplication|service="paymentTest
-```
+> unique metric = (metric_name + dimension_name + dimension_value)
 
+```python
+from aws_lambda_powertools.metrics import MetricUnit, single_metric
+
+with single_metric(name="ColdStart", unit=MetricUnit.Count, value=1) as metric:
+    metric.add_dimension(name="function_context", value="$LATEST")
+```
 
 ## Beta
 
+> **[Progress towards GA](https://github.com/awslabs/aws-lambda-powertools/projects/1)**
+
 This library may change its API/methods or environment variables as it receives feedback from customers. Currently looking for ideas in the following areas before making it stable:
 
 * **Should Tracer patch all possible imported libraries by default or only AWS SDKs?**
     - Patching all libraries may have a small performance penalty (~50ms) at cold start
     - Alternatively, we could patch only AWS SDK if available and to provide a param to patch multiple `Tracer(modules=("boto3", "requests"))` 
-* **Create a Tracer provider to support additional tracing**
-    - Either duck typing or ABC to allow additional tracing providers

python/aws_lambda_powertools/logging/logger.py

@@ -3,6 +3,7 @@
 import logging
 import os
 import random
+import warnings
 from distutils.util import strtobool
 from typing import Any, Callable, Dict
 
@@ -237,6 +238,7 @@ def log_metric(
         keyword arguments as additional dimensions (e.g. customer=customerId)
     """
 
+    warnings.warn(message="This method will be removed in GA; use Metrics instead", category=DeprecationWarning)
     logger.debug(f"Building new custom metric. Name: {name}, Unit: {unit}, Value: {value}, Dimensions: {dimensions}")
     service = os.getenv("POWERTOOLS_SERVICE_NAME") or service
     dimensions = __build_dimensions(**dimensions)

python/aws_lambda_powertools/metrics/__init__.py

@@ -0,0 +1,17 @@
+"""CloudWatch Embedded Metric Format utility
+"""
+from aws_lambda_powertools.helper.models import MetricUnit
+
+from .exceptions import MetricUnitError, MetricValueError, SchemaValidationError, UniqueNamespaceError
+from .metric import single_metric
+from .metrics import Metrics
+
+__all__ = [
+    "Metrics",
+    "single_metric",
+    "MetricUnit",
+    "MetricUnitError",
+    "SchemaValidationError",
+    "MetricValueError",
+    "UniqueNamespaceError",
+]