Bookmarks (4869)

Using load shedding to avoid overload On the surface, load shedding is simple. When a server approaches overload, it should start rejecting excess requests so that it can focus on the requests it decides to let in. The goal of load shedding is to keep latency low for the requests that the server decides to accept so that the service replies before the client...

How We Shipped PostgreSQL 14 on Azure Within One Day of its Release To our knowledge, this is the first time a major cloud provider has announced GA for a new Postgres major version on their platform one day after the official release. We’ll then describe the work involved in making Postgres extensions compatible with new major Postgres versions, including our distributed database Citus...

The insecure software supply chain: A history of failure and a new way forward supply chain attacks are more popular nowadays. What if NPM, Docker, Chef or any upstream sdk/provider are hacked and inject malicious code? Even a project as big as PHP had their Git server compromised. What can we learn and how can we prevent these.

When a JavaScript WeakMap() Actually Came in Handy Modern JavaScript has many low-level building block which looks like not useful at all. In this article we look into WeakMap and see why they are useful for our daily life of a front-end developer.

A Guide To CSS Debugging Debugging in CSS means figuring out what might be the problem when you have unexpected layout results. We’ll look at a few categories bugs often fit into, see how we can evaluate the situation, and explore techniques that help prevent these bugs.

xterm.js It’s a massive project to implement a terminal in browser. I don’t think I can ever read all of these but a few thing are useful such as see how they handle ascii color code.

fusuma multitouch gesture recognizer. This gem makes your linux able to recognize swipes or pinchs and assign commands to them. define your touchpad actions in yaml file and trigger command. Example, 3 left swipe, 4 up swipe etc.

Halo game's backend A team at Microsoft build a system for Halo 4 game. They talked about all scaling aspect, stress and load test and Vladimir was able to learn a lot from those public available resources. The system that they build is really impressive. They had to rework their legacy backend code, so they can achieve that their system can sustain an initial...

Herding elephants: Lessons learned from sharding Postgres at Notion They took Notion down for five minutes of scheduled maintenance. In that lone 5 minutes, they managed to cut over some Postgres upgrade which shard their database to help deal with the massive scale and leave room for future growth.

Scaling Large Production Clusters with Partitioned Synchronization This week’s paper review won a best paper award at Usenix ATC, and discusses Alibaba’s approach to scaling their production environment. In particular, the paper focuses on the evolution of the scheduling architecture used in Alibaba datacenters in response to growth in workloads and resourcesAn increase in resources or workloads impacted the load on the existing scheduler...

Path Building vs Path Verifying: The Chain of Pain On the last Saturday in May, at 10:48 GMT, a time when most folks in the US were still sleeping, the self-signed AddTrust External CA Root certificate expired. This is usually ok because those are usually well prepare ahead of time and the new root cert should be added in your local trust store. Then...

Gentle introduction to GPUs inner workings This article summarizes some lower level aspect of how GPU executes. Although GPU programming is not that complicated when compared to CPU, it also doesn’t match to what hardware is doing exactly. The reason is that we can’t just program GPU without some API, which is an abstraction over its inner workings.

Reverse Proxy, HTTP Keep-Alive Timeout, and sporadic HTTP 502s HTTP Keep-Alive between a reverse proxy and an upstream server combined with some misfortunate downstream- and upstream-side timeout settings can make clients receiving HTTP 502s from the proxy.

Why we spent the last month eliminating PostgreSQL subtransactions Gitlab journey to eliminates all SAVEPOINT call in their SQL queries that causes slow query, high cpu/disk uo, lock up connections. Fundamentally, the problem happens because a replica behaves differently from a primary when creating snapshots and checking for tuple visibility.

Zero-downtime Postgres schema migrations need this: lock_timeout and retries When you deploy database schema changes, you are not protected from system downtime even if you have very high-level automation but don’t use very low values of lock_timeout (or statement_timeout) to acquire a lock on the DB objects that are subject to change and do not implement some kind of retry logic. It’s better to...

Self-parking car in 500 lines of code We’ll train the car to do self-parking using a genetic algorithm in JavaScript. Hit the simulator to see how cool it’s