Build a rate limiting system for GenAI APIs. Learn token bucket implementation, quota management with Redis, and test multi-level limits with real load scenarios.
Learn practical strategies to prevent Denial of Wallet attacks in GenAI apps. Compare rate limiting algorithms and implement cost-aware protection with actionable checklists.
Prevent Denial of Wallet with cost-aware rate limiting. Learn how heterogeneous request costs shape design, trade-offs, and when to build custom limits.
This post defines the Staff Engineer as a senior IC who leads through influence and broad technical judgment. It offers high-level guidance on architecture, communication, learning, time management, and execution, and calls out common pitfalls to avoid.
Ownership in a services architecture extends beyond code. It includes the design, operational aspects and evolution of the service over time. This article explores the principles of code ownership in services, the team structures that support it, common pitfalls and as always - practical recommendations.
A short overview of AI tools for architecture-aware prototyping and knowledge-driven design workflows that we currently use and explore as software architects.
Exploring how the word "test" is often used in software development technique names in misleading ways, with examples from Test-Driven Development and Contract Testing.
Learn how to implement a scalable caching layer using Twemproxy, Memcached, and Ketama consistent hashing in Kubernetes. This article provides a practical guide for software architects and developers.
This article evaluates different caching architectures, comparing sidecar containers and dedicated caching proxies. It discusses the evolution of caching from client libraries to advanced proxy layers, highlighting their strengths and limitations in modern application architectures.
This post explores effective ways to manage high traffic and large-scale data by distributing caching responsibilities among multiple nodes. It emphasizes the significance of cache partitioning, contrasts it with data replication, and introduces essential methods like consistent hashing to ensure balanced load distribution.
In this blog post, we explore how embracing slow, analytical thinking, supported by Architecture Decision Records, can lead to more deliberate and durable architectural decisions.
In this short note, we share our thoughts on a panel discussion featuring Daniel Terhorst-North, who compared software development to surgery.
This blog post provides a comprehensive guide to effective planning for Site Reliability Engineering (SRE) projects, balancing long-term initiatives with operational responsibilities. However, the principles and practices discussed here are not limited to SRE. They can be applied to any software engineering project, making them relevant for a wide range of teams and organizations.
In this article, we’ll share our experiences with leveraging AI to efficiently generate high-quality Architecture Decision Records. We'll discuss practical techniques, provide examples, and outline the benefits and challenges we've encountered.
Recently, our team embarked on a major architectural transformation - moving from a monolith to a services-based architecture. As expected, the engineers on our team understood the rationale behind this shift, but the business stakeholders had concerns.
The Transactional Outbox Pattern is widely discussed in theory. In this blog post, we present a reference implementation of the pattern and provide a hands-on guide that bridges the gap between theory and real-world application.
After a short break from blogging, we are back with fresh insights from our recent presentation at the Poznań Java User Group about the practical DevOps approach.
As HandsOnArchitects marks its first anniversary, we're excited to revisit our top-performing blog posts that have resonated with our amazing community.
Knot.x serves as a pivotal integration layer, seamlessly blending content and style from Adobe Experience Manager with data from various sources to create sophisticated personalized websites, revolutionizing the construction and performance of digital platforms while offering flexibility and cost savings, making it a widely adopted solution in Adobe Experience Manager (AEM) projects.
In the intricate world of Kubernetes, the management and utilization of secrets within pods and containers demand careful consideration. In this article, we delve into available options, shedding light on how pods and containers are impacted across different scenarios.
In this post, we will share 5 practical tips and tools that help us manage time and keep our productivity high.
This post shows how to create a multi-component Helm chart and generate labels for each component
This blogpost presents an idea of using simplified environments for project demonstration purposes.
During my last presentation at AdaptTo() 2023, I mentioned reference content-centric architectures and how they leverage caching. In this blog post, I will expand on this topic.
This blogpost will present you an idea for a short, low-effort knowledge sharing sessions that we often use to spread out ideas and lessons learned among engineers.
This blog post is a short invitation for this year's AdaptTo session that Maciej Laskowski will be a speaker at.
In this brief note, we will share a simple way to migrate a website that uses Let's Encrypt certificate from one Kubernetes cluster to another with no downtime.
WebSight is an event streaming platform built by fully agile micro teams where every engineer is a technical lead.
This time we will show you a simple trick we found useful when setting up local or ephemeral k8s environments with a valid domain that supports multiple Ingress entries.
We will explain how the Pulsar client pre-fetch mechanism can impact the Pulsar learning curve.
Docker Desktop is a great tool that helps to save a lot of time during local development. However, it is not perfect and can break down in ways that are hard to restore, especially when you are using an M1 Apple silicon processor...
The Apache Software Foundation contribution guide to add Apple M1 | M2 laptops support for Docker
Technology Radar is a phenomenon that helps visualize the technology stack. Let's dive into how the tool may help to harness the tech stack at various levels of business.
Recently, we were discussing a post from Kent Beck on Code Reviews. Here are our thoughts.
This article describes the differences in message order when the recipient/consumer confirms or rejects messages subscribed to with an exclusive subscription.
