xeusnguyen.xyz

Search

SearchSearch

Recent Notes

See 149 more β†’

DueWeekly Tech: 24-03-2025 to 06-04-2025

Apr 06, 2025, 8 min read

center

Architecture

1. Blog - System Design was HARD until I Learned these 30 Concepts

  • The brief of popular concepts in technologies, if you want to prepare the interview, learn about bare concept and approach a new things what you want to explore about system design or system architecture, so this blog is one spent for y’all guys.
  • It sets the information, show case and explain why you should use once of these concept for your implementation.

2. Blog - Hardware-Aware Coding: CPU Architecture Concepts Every Developer Should Know (Insane Blog)

  • This blog cover about there popular of architecture of CPU, with cool illustrator and friendly explain
  • Author want to β€œAlong with exploration of these techniques, we will also look at multiple concrete code examples to understand how to apply this knowledge to optimize code performance”
  • Via Blog, Author want us to know more the how process work with your writing code, it doesn’t mean you write code with good behavior, it means you can control and understand how hardware align to process your high-level code. It’s tugh for sure.

3. Medium - 15 GitOps Best Practices to Improve Your Workflows

4. Medium - πŸš€ Master GitOps: Deploy Helm Charts with ArgoCD & Flux Like a Pro!

  • Article of Spacelift show us how to apply GitOps in good way and behavior, You choose GitOps for your project, product, so you should learn how to use this strategies in good ways. Don’t mess up, keep it clean and you can show the efficiency with GitOps.
  • Moreover, With GitOps and Kubernetes, you will work a lot with two tools and I think that truly popular idea than another candidate, Argo and Flux, both of them are great and have different way to use. But the second articles let us know about how combination with Argo and Flux to create wonderful solution, base infrastructure and developer behavior with Kubernetes Helm Chart.
  • Read more about feature of Argo and Flux, both of them have more things than CD only, double-check and you can find more what you need

5. Medium - Why Uber Moved from Postgres to MySQL

  • This articles show us how Uber deal to exchange from PostgreSQL to MySQL for resolving issue in Postgres17, index bloat, write amplification, replication overhead, data corruption risks, poor MVCC support on replicas, and difficult upgrades.
  • This articles take walkthrough for issue of PostgreSQL, like index bloat process via ctid and point to the MVCC make the new updated and that create problem, don’t have much experience but that’s kinda interesting.
  • In the end of articles, he try to let us know about how to mysql solve the problem of postgresql with the strategies in indexes.

6. Medium - Software Architecture for Developers

  • If you wanna find the pointer about your position with SA, you should read that one, cuz it give you a couple of properties of SA attached with Developers
  • You will find how to growth up your skill about SA via this one, I means it can help you understand what high level compare with your level about thinking arch, that’s really good and opportunity for learning

AI with Kubernetes

1. Medium - Deploying Ray Cluster for AI/ML workloads on a Kubernetes Cluster

2. Medium - Effective Load Balancing with Ray on Amazon SageMaker

3. Alibaba - Best Practices for Ray Clusters - Ray on ACK

  • If you approach with AI and Machine Learning in Kubernetes Environment or Local Developer, You will meet the Ray - The framework help you scale and more distribute the AI Solution in multiple platform, and KubeRay is the thing created for this responsibility
  • These articles show us how we can use and operate best practice for Ray in Kubernetes, SageMaker and we can deal with them to create the best behavior when you deploy Ray Cluster into environment, and get the good value to manage them. Explore more about Ray at Documentation
  • Ray is a great techstack but it’s really complicated, you need to make sure to understand this native way to get the good approach and behavior.
  • More alternative for Ray, such as MLflow, Spark and Dask

Data Engineer

1. Medium - Setting up a Spark standalone cluster on Docker in layman terms

2. Medium - Setting Up Apache Spark from Scratch in a Docker Container: A Step-by-Step Guide

3. Medium - Apache Spark & Airflow in Docker: Step by Step guide

  • I just give time to read about Spark and here are what things you can need to self-hosted and playground with Spark, In docker. If you wanna find how to set up Spark in Kubernetes, You should consider to find Spark Operator to help you set up Spark in your Kubernetes environment
  • The things I like most of these blogs because they write about set-up guide with very detail and easier for approaching, you can read and follow with their strategies and source code attached - It’s truly useful.
  • In the last, I see the cool stories about how can we combine Spark and Airflow, that’s really interesting question. You can double-check about to learn how they structure Spark with Airflow, may be you can use both of them for curial purpose. E.g: You try to run Apache Beam for building your pipeline with remote runner, like Spark, Flink, or Google Data Flow. If you wanna try Beam with Spark, I think this article Medium - How to run Apache beam on Spark cluster is good for y’all. NOTE (Remember Java 11 is good start for Beam and Spark, Trust me)

Story

1. Blog - What High Performers Know About Doing Hard Things

  • Awesome blog talk about the hard things can bring you more value than you think, the pressure and discomfort are the chance for your growth up
  • The author give you advise deal with problems, deal with hard things, deadlines, delay and whatever come up in your career, especially your techniques skill, mindset and habit, those reasons are the best things can convince yourself stand again and do more hard things.
  • It’s pretty inspiration for me.

Technologies

1. Youtube - Configure OIDC access to Vault in Less than 10 Minutes!

2. Medium - Vault + Keycloak: Secure JWT-Based Access

3. Vault - Use JWT/OIDC authentication

  • Super detail tutorial and manual for who want to configure oidc/jwt authentication for Vault
  • It supposes and recommends for you about strategies, what action should do and what command line should trigger for whole progress work
  • I prefer it so well because It’s newest version and can help you imagine with Vault in 2025, BTW you can use another resources but you should follow with this useful documentation

4. Medium - Paying for software is stupid: Open-Source tools to Destroy Your SaaS Expenses

  • The collections of massive tools for multiple purpose from AI, Data, Project Management, Workflow, … and some candidates actually work as good result like, n8n, airbyte, superbase, …
  • He try to introduce the best feature of these tools for us to find and explore which compatible for us, and it’s kinda useful for some meaning πŸ™Œ

5. Medium - 7 Bash Scripts for Quick Troubleshooting in Production Environments

  • Couple of scripts are useful for helping debug linux machine, and It pretty well for us to quick check about performance issue inside your server
  • You can covert that into bunch of script, and reuse at time you actually want it. Honestly, if you are DevOps Engineer or System Administrator, you should go for it.

6. Medium - Apache Kafka 4.0: What’s New and Why This Upgrade is a Total Game-Changer

7. Medium - Strimzi β€” Deploy Kafka in Kubernetes

8. Strimzi - Migrate your Strimzi-operated cluster from ZooKeeper to KRaft

  • These articles cover a lot about what new features of Kafka Version 4.0.0, Which replace zookeeper - legacy services integrated with Kafka by Kraft. If you wanna explore more about Kraft and zookeeper, and what change of them with Kafka, you should visit Medium - The Evolution of Kafka Architecture: From ZooKeeper to KRaft
  • In these articles, I just read about how to use strimzi - Kafka Operator for Kubernetes - This platform will help us deploy Kafka Cluster with less effort and easier configuration, you can do more things with this operator to create expectation Kafka Cluster for yourself in Kubernetes environment
  • Moreover, If you already use Zookeeper with Kafka, Strimzi will show us the way to migration from Zookeeper to Kraft, which directly and manage itself by Kafka, It’s wonderful than you need to manage Zookeeper.

Graph View

Backlinks