How to Build Production Observability with Python, Prometheus and Grafana

·1 views

Modern distributed systems demand observability — understanding why and where failures occur — rather than simple uptime monitoring. A practical approach involves collecting three types of telemetry: metrics, logs, and traces, with metrics best structured around the RED method covering request rate, errors, and duration. Developers can instrument a Python FastAPI service using the prometheus_client library to expose custom metrics at a dedicated endpoint for Prometheus to scrape. The full observability stack, including Prometheus for data collection and Grafana for visualization, can be run locally via Docker Compose with minimal configuration. This setup enables engineering teams to detect, diagnose, and resolve production issues faster without relying on manual log searches.

Read the full story at DEV Community

This is an AI-generated summary. ShortSingh links to the original source for the complete article.

Discussion (0)

Tutorial: Build a Local Services Flutter App Using HosteDay Backend Platform

A new developer tutorial demonstrates how to build a local services app called 'At Your Service' using Flutter and the HosteDay backend platform. The guide walks through creating a free backend server on HosteDay, setting up a database table for service providers, and auto-generating CRUD API endpoints without writing manual backend code. Developers then create a Flutter project in Android Studio and integrate the hosteday_flutter package to connect the app to the backend. API token protection is configured within the Flutter project to secure backend requests from unauthorized access. This first part of the series lays the foundation for future authentication and service-listing features.

0 comments Read more at DEV Community

ProgrammingDEV Community ·

How HTTP/2 Multiplexing Replaced Multiple Connections with One Efficient Link

HTTP/1.1 handled growing web traffic by opening multiple TCP connections simultaneously, which improved page load speeds but created significant overhead in handshakes, memory, and CPU usage. Engineers identified that the core issue was not network capacity but how HTTP utilized the network. HTTP/2 was developed to address this by introducing multiplexing, which allows multiple requests and responses to travel concurrently over a single TCP connection as independent streams. The protocol breaks data into smaller units called frames, which can be interleaved across streams and reassembled correctly by the browser. HTTP/2 also shifted from a text-based format to binary, making it more efficient for computers to process.

0 comments Read more at DEV Community

ProgrammingDEV Community ·

Guide: Migrating Python Projects from mypy to ty and Pyrefly

A practical migration guide documents how the fastkml Python package was moved from mypy to Astral's ty and Meta's pyrefly type checkers. The guide recommends running both tools simultaneously, as they each catch different subsets of type errors that the other may miss. Before making any changes, developers are advised to establish a baseline error count and categorize issues by error kind to identify systemic root causes rather than fixing problems file by file. The guide also warns that mypy configuration flags do not map one-to-one to ty or pyrefly equivalents, urging developers to re-derive intent rather than hunt for identical settings. Finally, it stresses verifying that both tools pass cleanly alongside a full test suite before retiring the existing mypy configuration.

0 comments Read more at DEV Community

ProgrammingDEV Community ·

Free PHP Scripts on Forums Found Stealing Server Credentials and Sensitive Data

A developer discovered that a free PHP script downloaded from platforms like 4shared and online forums contained hidden malicious code. The script silently collected and transmitted sensitive server data — including email credentials, usernames, passwords, and environment details — to an external destination without the user's knowledge. The malicious code used obfuscation techniques such as base64 encoding and dynamic execution functions to disguise its true purpose and evade casual inspection. The incident highlights the risks of using unverified third-party code, particularly free scripts shared outside reputable platforms. Security experts recommend auditing all third-party code for suspicious functions like eval() and base64_decode() before deploying it in any production environment.

0 comments Read more at DEV Community