📖
Paper Notes
Selected Paper Interpretation Notes.
Skip to content 
Blog Articles
Covers five major areas: database principles, framework tools, programming languages, algorithm analysis, and model analysis.
Database and system tuning
From database core principles to performance optimization practices, help developers build highly available data systems.
Framework tool guide
In-depth analysis of the core mechanisms of popular frameworks and tools such as Redis and Kafka, and release operation guides and best practices.
Programming language analysis
Focus on the underlying principles of languages such as Go/Python to improve code writing efficiency.
Algorithm analysis
Systematically explain the implementation and application scenarios of classic algorithms such as Raft consensus algorithm, genetic algorithm, and Bloom filter.
Model interpretation
Interpret the architecture design principles of machine learning models such as large models, combine mathematical derivation with engineering practice cases, and strengthen the algorithm implementation capabilities.
Series
Focusing on paper interpretation, data mining and programming design, it helps developers advance their learning with systematic and in-depth content.
Paper and Data Mining Practice Column
Selected top computer conference papers and authoritative works in the industry for reading and analysis.
Programming Design Guide
In-depth interpretation of the seven core principles of object-oriented programming, combined with code examples to analyze the design logic of high cohesion and low coupling, provide practical demonstrations of scalable architecture, and help developers build easy-to-maintain and highly reusable software systems.
How to safely store user passwords in the backend? Add salt!
Almost all systems require users to register and log in, and the most common way to log in is to let users enter their username and password. Many early systems only required usernames and passwords to log in, but some websites have been attacked by brute force, social engineering refers to an attack method that manipulates human psychology rather than technical means to obtain passwords or sensitive information. The core is to exploit human weaknesses, such as trust, fear, curiosity or negligence, to induce the target to actively leak information or perform specific operations. refers to the data in the database being directly exported by the attacker.
How to reduce inventory under high concurrency?
This is a common scenario in e-commerce. Currently, Internet giants have very mature solutions for it. I wrote this article just to give some of my own thoughts.
MVCC Mechanism of MySQL Database
MySQL, as a multi-threaded database, supports concurrent queries from clients and sets its default isolation level to repeatable read. So how does MySQL isolate transactions in concurrent operations? It actually uses the
What is HDFS?
If you are also a big data worker, you must have heard of the distributed file system:
Overview of Fundation Models
With the development of the times, the application of foundation models in various fields is expanding. This article tries its best to sort out various materials and will give a brief overview of foundation models from the aspects of concept definition, type classification, training and application.
About Redis Persistence And Clustering
Redis is a very commonly used key-value database. The use of memory and HashMap for storage brings efficient queries. This article will introduce the persistence principle of Redis, the deployment method of clusters, and the expiration and elimination mechanism of key-value pairs.
What problems will we encounter in the process of ensuring the consistency of distributed transactions?
In today's microservice architecture, how to ensure the consistency of distributed transactions is a problem that every backend development engineer may encounter.
Why is the number of distributed nodes usually an odd number?
As a backend engineer, I often deal with various distributed systems in my daily work, such as ETCD, Redis, k8s, etc. When deploying these distributed clusters, we usually set the number of nodes to an odd number, which seems to be a conventional rule. But why? In addition to the fact that even-numbered nodes are prone to voting tie, are there other reasons?
ORB-SLAM2 Installation Guide
This blog will describe in detail how I installed Ubuntu16.04 dual system in bare metal state and installed the corresponding version of ROSkinetic system. In addition, after installing these two systems, how to successfully install and debug INDEMIND binocular camera and implement ORB-SLAM2 algorithm.
Keep It Simple
Everything can be explained in one sentence.
ShenZhen, China
Announcement: Limited English Blog Content
Dear readers,
Thank you for visiting my blog! I want to acknowledge that there aren't many posts in English here at the moment. As a native Chinese speaker, I've always written primarily in Chinese. However, I'm now working to translate my existing content into English to reach more readers.
This process involves using tools like Google Translate, so you may encounter occasional less-than-perfect expressions. I truly appreciate your understanding and patience as I navigate this transition.
The translation effort will take time, but I'm committed to bringing more content to you in English. Thank you for your support and for sticking around! 😄