Unfazed Prometheus

Installation

pip install unfazed-prometheus

Quick Start

Add Settings

# settings.py
import socket

UNFAZED_PROMETHEUS_SETTINGS = {
    "HOSTNAME": socket.gethostname(),
    "PROJECT": "{{ project_name }}",
    "PROMETHEUS_MULTIPROC_DIR": "/prometheus",
}


# add lifespan

UNFAZED_SETTINGS = {
    # ... other settings
    "LIFESPAN": ["unfazed_prometheus.lifespan.PrometheusLifespan"],
}

Monitor Request

all you need to do is add the middleware to the middleware list.

UNFAZED_SETTINGS = {
    # ... other settings
    "MIDDLEWARE": [
        "unfazed_prometheus.middleware.common.PrometheusHttpRequestMiddleware",
    ],
}

Monitor Database using Tortoise ORM

all you need to do is to use the unfazed_prometheus.database.tortoise.mysql database engine.

UNFAZED_SETTINGS = {
    "DATABASE": {
        "CONNECTIONS": {
            "default": {
                "ENGINE": "unfazed_prometheus.database.tortoise.mysql",
                "CREDENTIALS": {
                    "HOST": "mysql",
                    "PORT": 3306,
                    "USER": "app",
                    "PASSWORD": "app",
                    "DATABASE": "app",
                },
            }
        },
    },


}

Monitor Cache

all you need to do is to use the unfazed_prometheus.cache.backends.default.PrometheusDefaultBackend cache backend.

other cache backend:

• unfazed_prometheus.cache.backends.namespace.PrometheusNamespaceBackend
• unfazed_prometheus.cache.backends.serializer.PrometheusSerializerBackend

UNFAZED_SETTINGS = {
    "CACHE": {
        "default": {
            "BACKEND": "unfazed_prometheus.cache.backends.default.PrometheusDefaultBackend",
            "LOCATION": "redis://redis:6379",
            "OPTIONS": {
                "decode_responses": True,
                "max_connections": 1000,
            },
        }
    },
}

Monitor Function

use prometheus agent.monitor_function decorator.

from unfazed_prometheus import agent


@agent.monitor_function
def my_function(a: int, b: int) -> int:
    return a + b

Monitor API

use prometheus agent.monitor_api decorator.

from unfazed_prometheus import agent


@agent.monitor_api("/api/v1/users")
async def get_users():
    resp = await asyncrequests.get("https://api.github.com/users")
    return resp.json()

Monitor Exception

unfazed_prometheus will monitor exceptions through other monitor decorators. all you need to do is to let the exception be raised or raise it yourself.

from module import CustomException

@agent.monitor_function
def devide(a: int, b: int) -> float:
    if b == 0:
        raise CustomException("b is 0")

    return a / b

or just let the exception be raised.

@agent.monitor_function
def devide(a: int, b: int) -> float:
    return a / b

unfazed_prometheus will automatically collect the exception and count the total number of exceptions.

connect to prometheus server

in the live env, it's better to connect to prometheus server use another service other than the unfazed server.

example code see scripts/prometheus_client.py

Advanced

if the default metrics and decorators cannot meet your needs, meta_monitor may help.

from prometheus_client import Counter, Histogram
from unfazed_prometheus import meta_monitor


counter = Counter(
    "my_counter",
    "my_counter_description",
    ["label1", "label2"],
)


my_monitor = meta_monitor(
    counter_handler=counter,
    counter_labels=["foo", "bar"],
)


@my_monitor
def my_function(a: int, b: int) -> int:
    return a + b

meta_monitor signature:

def meta_monitor(
    counter_handler: t.Optional[Counter] = None,
    hist_handler: t.Optional[Histogram] = None,
    exc_handler: t.Optional[Counter] = None,
    counter_labels: t.Optional[Labels] = None,
    hist_labels: t.Optional[Labels] = None,
    exc_labels: t.Optional[Labels] = None,
) -> Decorator:

    ...