标签：#orm

Imagine a user trying to upload a 100MB video or a high-resolution photo to your app. If you use the standard Rails file upload, that file travels from the user's browser to your Rails server, and then your server sends it to S3 or Google Cloud. This is a terrible way to do it. While that 100MB file is transferring, your Rails worker (Puma) is frozen. It can't handle other users. If three people upload large files at once, your whole app will stop responding. In 2026, the professional way to handle this is Direct Uploads . With Direct Uploads, the file goes directly from the user's browser to your cloud storage (S3, R2, etc.). Your Rails server only handles a tiny bit of metadata. It is faster for the user and much safer for your server. Here is how to set it up in Rails 8. STEP 1: Configure Your Storage First, make sure you aren't using the local disk for production. You need a cloud provider like AWS S3 or Cloudflare R2. In your config/storage.yml : amazon : service : S3 access_key_id : <%= ENV['AWS_ACCESS_KEY_ID'] %> secret_access_key : <%= ENV['AWS_SECRET_ACCESS_KEY'] %> region : us-east-1 bucket : my-app-uploads # Crucial for Direct Uploads! public : true Note: You must configure CORS in your S3/R2 dashboard to allow requests from your domain. If you don't do this, the browser will block the upload. STEP 2: The Rails Form Rails makes the backend part incredibly easy. You just add one attribute to your file field: direct_upload: true . <!-- app/views/users/_form.html.erb --> <%= form_with ( model: user ) do | f | %> <div class= "field" > <%= f . label :avatar %> <%= f . file_field :avatar , direct_upload: true %> </div> <%= f . submit "Save Profile" %> <% end %> When you add direct_upload: true , Rails automatically includes a JavaScript library that handles the "handshake" with S3. STEP 3: Adding a Progress Bar (The UX Win) Direct uploads can take a few seconds. If nothing happens on the screen, the user will think your app is broken. We can use the built-in Ac

Logging has a habit of ending up in the places you care about most. It starts as a few lines for visibility. Then those lines appear in request handling, market-data processing, matching loops, telemetry pipelines, and other code where predictable latency matters. At that point, a log statement is no longer just observability. It is work running on the same thread you are trying to keep fast. A line like this can look harmless: LOG_INFO ( logger , "order_id={} price={}" , order_id , price ); The important question is what happens before the caller continues . Does it evaluate expensive arguments? Format text? Copy buffers? Allocate? Contend with other producer threads? Wait for queue space? For many applications, those costs are acceptable. For latency-sensitive systems, they are part of the latency budget . spdlog is one of the best-known C++ logging libraries and a strong general-purpose choice. It is mature, easy to use, and has a broad feature set. Quill was designed for a narrower problem: How little work can a C++ logger leave on the caller thread while still producing rich, human-readable logs? That is the lens for this comparison. The interesting difference is not which library has more features. It is where each library chooses to spend work. At a Glance Area spdlog async Quill User-message formatting Producer thread Backend thread Producer handoff Shared thread-pool queue Per-thread SPSC queue Arguments for runtime-disabled levels Evaluated if the level was not compiled out Skipped by the macro-level runtime check Native synchronous mode Yes No Backend workers Configurable thread pool Single backend worker Primary focus General-purpose flexibility Low producer-side latency These differences do not make one library universally better. They make each library better suited to different workloads. Async Logging Is Not One Design "Async logging" often means "file I/O happens on another thread." That is useful, but it is not enough to describe the cost paid by t

For the past four years, a sprawling Android-based botnet called Popa has forced millions of consumer TV boxes to relay Internet traffic linked to advertising fraud, account takeovers, and mass data-scraping efforts. This week, researchers from multiple security firms concluded that the Popa botnet is linked to NetNut, a "residential proxy" provider operated by the publicly-traded Israeli firm Alarum Technologies Ltd [NASDAQ: ALAR].

Vinay Chella and Akshat Goel discuss the challenges of running traditional CDC across heterogeneous databases during peak order traffic. They explain how Debezium hit limits under high load and share how they built Write-Ahead Intent Log (WAIL) - a custom architecture that utilizes a dumb producer proxy and a smart consumer pattern to cleanly separate the intent from the state payload. By Vinay Chella, Akshat Goel

I shipped a small dev-news reader to Google Play with the entire client written in one Compose Multiplatform codebase — every screen in commonMain , no per-platform UI. This is an honest field report on what that actually costs in production: the numbers, the native seams, and the parts that still hurt (hi, iOS). The repo is open source (MIT), so everything here is checkable. TL;DR — For a content/list/detail app, CMP is comfortably production-ready on Android. 96.9% of the shared module is commonMain ; the native cost is concentrated in ~10 expect/actual seams. iOS compiles and renders, but isn't polished yet. The numbers Source set Files Lines Share commonMain 59 ~7,700 96.9% androidMain 2 123 1.6% iosMain 3 127 1.6% All 17 screens live in commonMain — trending list, aggregated feed, README detail, profile with paging, settings, favorites — and there isn't a single if (isAndroid) branch in the UI. The 3% that isn't shared The native cost isn't spread thinly across the codebase. It's concentrated in ~10 expect/actual seams, and this is the entire list: Platform info — app version, system language, User-Agent string System interaction — open URL, open app settings, share sheet Analytics — a trackEvent hook (Android → Aptabase; iOS is deliberately a no-op for now) WebView — the messy one (below) Everything that touches a platform API is small and enumerable. Everything else came for free. Why expect/actual and not an interface + DI? For these ~10 seams, expect/actual was the least ceremony: no DI wiring, and the compiler refuses to build until every target implements the declaration. The moment a seam has more than one implementation, or I'd want to fake it in tests, an interface in commonMain with injected impls is the better tool. For a fixed set of platform primitives, expect/actual wins on friction. The ugliest boundary: WebView I have two WebView paths, and I'll be precise because the repo is open: Rendering GitHub READMEs from an HTML string — inline expect/act

C++ source code is written in order. That does not mean the processor executes it in order. This is the first correction. It is also the one many performance discussions manage to avoid. Modern high-performance cores use out-of-order execution . They accept a sequential instruction stream, break it into internal operations, rename registers, place work into scheduling structures, execute ready operations early, and then retire the results in program order. The machine preserves the visible behavior of sequential execution. Internally, it is not taking attendance line by line. For ordinary software, this is mostly invisible. For C++ intended to run in tens of nanoseconds, it is not invisible. At that scale, performance is not just about the number of instructions. It is about whether those instructions can be scheduled in parallel or whether the program quietly built a dependency chain and then acted surprised. The processor is a dependency scheduler Out-of-order execution exists because in-order pipelines waste time. If an older instruction stalls, an in-order processor must often wait even if later instructions are independent and ready. That is a poor use of hardware. The chip has execution units available. The instruction stream has more work. Dynamic scheduling fixes part of this problem. The processor tracks which operations have their inputs ready. When an operation is ready and an execution unit is available, it can issue. Older operations may still be waiting. Later operations may run first. The final architectural state is still committed in order, so the program behaves correctly. Tomasulo’s algorithm is the classic model for this idea. It used reservation stations and register renaming to allow instructions to execute when their operands became available rather than strictly when they appeared in the original program (Tomasulo, 1967). Later superscalar processors extended the same general approach with speculation and reorder buffers, but the central idea

How GPU scheduling complexity and MLOps integration are forcing platform teams to rearchitect Kubernetes clusters before operational debt becomes insurmountable. As AI workloads consume roughly 40% of enterprise Kubernetes clusters by 2026, the platform's default scheduler is proving fundamentally mismatched with the topology-aware, gang-scheduled demands of GPU-intensive training and inference. Platform engineering teams that invest now in purpose-built GPU scheduling layers, multi-tenant partitioning, and FinOps-driven autoscaling will separate themselves from organizations drowning in 30-45% GPU utilization rates and mounting infrastructure costs. Why the Default Kubernetes Scheduler Fails GPU Workloads Kubernetes was designed for stateless, CPU-bound services, and its pod-by-pod bin-packing scheduler has no native awareness of GPU topology, NUMA boundaries, or NVLink interconnect bandwidth. This becomes a critical failure point with NVIDIA H100 SXM5 nodes, where achieving full-bandwidth tensor parallelism requires all 8 GPUs on a node to be scheduled as a single atomic unit. The default scheduler cannot guarantee this co-placement, meaning distributed PyTorch FSDP or MPI training jobs frequently land on suboptimal node configurations, wasting expensive NVLink bandwidth and forcing teams to over-provision GPU capacity. Idle GPU memory stranded across partially-utilized nodes is the primary driver behind the 30-45% utilization rates reported in 2025 surveys by Gradient Dissent and Weights and Biases, representing millions of dollars in annual wasted spend for mid-to-large enterprises running mixed AI workloads. Building the GPU Scheduling Stack: Volcano, KAI Scheduler, and MIG Platform teams are converging on a layered scheduling architecture that replaces or augments the default Kubernetes scheduler with GPU-aware primitives. Volcano has become the dominant choice for distributed training workloads, using its PodGroup abstraction to enforce gang scheduling across

Clair Health will track inflammation and bloating markers, energy levels, and cycle phase classification to give insights into cycle irregularities and perimenopause, as well as hormonal fluctuations, and how to navigate those changes.

Pramaana will focus on highly sensitive verticals like law, drug discovery, and tax preparation — where errors can be costly and reliability is at a premium.

TL;DR — UUID v7 trie 13× plus vite que v4 en simulation B-tree (1M entrées), expose une empreinte mémoire identique, mais révèle son timestamp d'émission. UUID v4 reste le choix "zéro réflexion" pour les identifiants isolés. Le reste de cet article vous donnera les données pour décider. Introduction Les UUIDs sont omniprésents dans les systèmes modernes : clés primaires de bases de données, identifiants de sessions, tokens de traçabilité. Pourtant, le choix de la version impacte directement les performances en écriture et en lecture, la fragmentation des index, et — dans certains contextes — la confidentialité des données. UUID v4 (RFC 4122, 2005) est aujourd'hui la version par défaut de presque tous les ORM et frameworks. UUID v7 (RFC 9562, 2024) est son successeur moderne, conçu pour corriger son principal défaut : le désordre lexicographique. Dans cet article, nous allons mettre les deux en face avec des benchmarks réels sur des volumes de 100 000 à 10 millions d'UUIDs , analyser leur structure bit par bit, et vous donner une grille de décision claire. Structure interne : ce que contiennent ces 128 bits UUID v4 xxxxxxxx-xxxx-4xxx-yxxx-xxxxxxxxxxxx Bits Contenu 0–47 Aléatoire 48–51 Version (0100 = 4) 52–63 Aléatoire 64–65 Variant (10) 66–127 Aléatoire 122 bits d'entropie pure. Aucune information temporelle. Chaque UUID est statistiquement indépendant des autres. UUID v7 019ed5c8-2a2f-7974-91f2-6ba1f313dcfa └──────────────┘ 48 bits = timestamp Unix en millisecondes Bits Contenu 0–47 Timestamp Unix (ms) 48–51 Version (0111 = 7) 52–63 Aléatoire (sub-ms ou compteur) 64–65 Variant (10) 66–127 Aléatoire 74 bits d'entropie + 48 bits de temps. Naturellement monotone : deux UUIDs générés dans la même milliseconde sont toujours distincts et ordonnés de façon cohérente. Lecture du timestamp (Python) : import uuid6 u = uuid6 . uuid7 () b = u . bytes ts_ms = int . from_bytes ( b [: 6 ], ' big ' ) # → 1781703125553 (ms depuis epoch Unix) Depuis la sortie de nos benchmarks : [00

IaC tools built for single-team deployments fail structurally at 200 accounts because the failure modes are architectural, not configurational. Why 200 Accounts Is Where IaC Tools Break IaC tools built for single-team deployments fail structurally at 200 accounts because the failure modes are architectural, not configurational. Scale Threshold State Management Provider Auth Overhead Execution Time Impact ~10 accounts One backend bucket, one workspace; quirks are routable Not a meaningful bottleneck Sequential plan/apply is manageable ~50 accounts Sequential execution still viable; blast radius contained Per-account latency exists but tolerable Below threshold where parallelism is required 200+ accounts 200 separate plan operations per shared-module refactor 3-sec per-account auth × 200 = 10 min added per plan cycle Sequential Terraform applies measured at 4.1 hours end-to-end A 10-account environment forgives sloppy state management. One backend bucket, one workspace convention, one pipeline. Engineers learn the tool's quirks and route around them. At 200 accounts, those same quirks compound. Why the threshold is 200 State lock contention, cross-account provider authentication chains, and module resolution latency stack on top of each other. The result is not slower deploys. It is non-deterministic deploys, which is operationally worse. The specific threshold matters. Below roughly 50 accounts, most teams run plan and apply sequentially without parallelism because the blast radius of a runaway apply is contained. Above 200 accounts, sequential execution becomes untenable. We measured a 200-account org where sequential Terraform applies across all accounts took 4.1 hours end-to-end. That latency made emergency remediation impossible inside a standard incident window. Three compounding failure modes State file proliferation. Each account carries its own state file, and each state file is a consistency boundary. At 200 accounts, a single refactor touching a shared modu

As a backend developer, I sometimes help companies evaluate candidates by reviewing their recorded technical interviews. However, over time, I’ve noticed a deeply ingrained misconception. When discussing memory management or data streaming, many developers explicitly state: "Iterators in Python are inherently eager. If you want true lazy loading or lazy evaluation, you have to use generators and the yield keyword." This misconception is common. Many popular bootcamps and online courses introduce lazy evaluation exclusively through generators . Custom class-based iterators are usually skipped or dismissed as boilerplate-heavy OOP theory rarely used in production Python. This confusion is further reinforced by two common educational simplifications : The List vs. Generator Expression Analogy: Beginners are taught that square brackets [...] (list comprehensions) are eager and take up memory, while parentheses (...) (generator expressions) are lazy. This often creates a false binary mental model: "generators = lazy, everything else = eager." Standard "Textbook" Examples: When courses demonstrate a custom iterator, they usually write a basic class that accepts an already fully loaded list in its __init__ and simply increments an index in __next__ . While this is valid for in-memory data, it leads developers to assume that custom iterators inherently require loading all data upfront. In reality, generators are a specialized language feature designed to implement the iterator protocol automatically . They comply with the exact same interface ( __iter__ and __next__ ). A generator is lazy not because of some magical property of the yield keyword, but simply because it adheres to this underlying contract. To show that custom iterators can be lazy without using any generators or yield keywords, I’ve put together a lightweight and reproducible benchmark. 🧪 The Experiment: Proving Lazy Loading with Custom Iterators Suppose we need to read a database export file ( test_users_db.

The problem IAM wildcards and public S3 buckets keep slipping through Terraform code review. Tools like Checkov and tfsec exist but they live in CI, require config files, and developers ignore the output because it's not where they're working. What I built TerraWatch is a GitHub App that scans every pull request that touches .tf files automatically. If it finds a security issue it blocks the merge and posts the exact code fix as a PR comment. The developer sees something like this in their PR: ⚠️ PUBLIC_S3_BUCKET - main.tf (Line 6) Severity: HIGH Risk: S3 bucket allows public read access. Fix: acl = "public-read" acl = "private" block_public_acls = true restrict_public_buckets = true They copy the fix, push, and the merge unblocks automatically. How it's different No YAML, no CI config - installs in 2 minutes via GitHub App Fixes are hardcoded diffs, not AI generated Nothing auto-applied - you review every fix No Checkov dependency - own lightweight rules engine Only reads changed .tf files in the PR, never your full codebase 29 rules covering S3 public access, IAM wildcards, open ports (SSH/RDP/MySQL/Postgres), unencrypted EBS/RDS, public databases, hardcoded secrets, EKS public endpoints, CloudTrail disabled, IMDSv1, and more. Try it Free during beta - terrawatch.dev Also launching on Product Hunt today if you want to show some support!

A bot on Polymarket quietly extracted $32k in near risk-free profits by sniping “Will Company XYZ Beat Earnings?” markets. It waits for the official release, then instantly buys the winning side. Many limit orders from retail traders remain uncancelled, creating a post-announcement arbitrage window. Two developers decided to challenge it. Here’s what they learned while trying to build a faster version. Infrastructure Choices Location : Polymarket’s CLOB runs in AWS eu-west-2 (London). They deployed from Ireland (eu-west-1, Dublin) — the closest realistic option without IP tricks. UK IPs are blocked. Language : Rust for type safety and speed. The author notes you can achieve competitive latency in Python if you strip unnecessary network calls. Key Warning : Avoid the official Polymarket SDKs for ultra-low latency. They include helpful but slow pre-trade checks. Build lean custom clients. The Data Feed Challenge (The Real Bottleneck) The critical edge is getting earnings announcements faster than competitors. Source Performance Verdict Scraping Newswires Too slow Failed Benzinga Low-Latency Slower than manual clicking Failed Paid ultrafast feed ~500ms after release Still too slow EDGAR Consistently slower than newswires Backup only Even at 500ms, the order book was already swept by faster bots. The top players are likely using extremely expensive dedicated feeds or custom setups. Technical Lessons Learned Network > Code Most latency lives in the network round-trip, not in language choice. Optimize transport first. Custom Execution Layer Skip heavy SDK abstractions. Direct signed orders with minimal validation. Post-Event Sniping Logic Monitor newswire feeds aggressively Parse EPS vs. estimate instantly Place aggressive limit/market orders on the winning side Handle cases with ambiguity (multiple interpretations of “beat”) Reality Check They made some wins during EPS ambiguity or when faster bots hit size limits, but never won on pure speed against the leader. Why This

Armed with a ton of new upgrades, Ferrari came to Spain full of confidence.

Pipeline & Prompts | Byte size guides on DevOps, Cloud and AI ** AI in the Stack #1** Byte size summary After reading this article, you'll have a framework for evaluating AI tools in platform engineering contexts — not by capability type, but by where in your workflow the tool actually changes the outcome. You'll understand why the tools that sound most compelling are still hype, where genuine productivity gains exist today, and what governance infrastructure you need in place before any AI component gets near production. This article is the foundation for the series; subsequent articles implement each touch point against real OpenShift infrastructure. The story I spent months selling IBM's AI and data science portfolio before I truly understood what I was selling. I knew the pitch. Predictive analytics. Optimization. Decision intelligence. I could walk a room through the business value without breaking a sweat. CPLEX for scheduling, Watson for insights — I had the slides, the talking points, the customer stories. Then I sat in on a data scientist demo. Not a sales demo. An actual working session — models being trained, outputs being interrogated, assumptions being challenged in real time. And somewhere in that room, watching someone do the thing I'd been describing from the outside, something clicked — and not in a good way. The models were impressive. The theory was solid. But I kept asking myself the same quiet question: where does this go next? Because most of what I saw never made it anywhere near production. It lived in notebooks. In slide decks. In proof-of-concept environments that were never ready to cross the line into something real. I'd been selling outcomes — optimised schedules, smarter decisions, reduced costs — without a clear path to how you'd actually get there. And underneath all of it, something else bothered me that nobody was talking about loudly enough: the data going into these models was often messy, unvalidated, and ungoverned. Bias wasn't

Pipeline & Prompts | Byte size guides on DevOps, Cloud and AI AI in the Stack #2 ⚡ Byte Size Summary RAG inserts a retrieval layer between your existing runbooks and an LLM — answers come from your documentation, not generic training data, with source citations included. This article builds a complete FastAPI service with /ingest , /query , and /health endpoints, using OpenAI embeddings and Chroma as the vector store. Everything is cloneable from GitHub. The goal is not to replace your runbooks. It is to make them queryable at the moment an incident is happening. I have never met a platform team with bad runbooks. I have met plenty of platform teams where the runbooks exist, are reasonably well written, are stored somewhere sensible — and are still completely useless at 2am when something is on fire. Not because the content is wrong. Because nobody can find the right one fast enough. The search in Confluence returns fourteen results and none of them are titled the way the engineer is thinking about the problem. The person on call is junior and doesn't know the runbook exists. The runbook was written for a slightly different version of the service and nobody updated it. The runbook problem is not a writing problem. It is a retrieval problem. That is exactly the problem RAG was built to solve — and it is one of the highest-ROI first applications of AI in a platform engineering context. Not because it is technically impressive. Because it closes a gap that costs your team hours every month. This article builds a working pipeline. By the end you will have a FastAPI service that takes a natural language question — "why is my pod stuck in CrashLoopBackOff after a config change?" — and returns an answer grounded in your actual runbooks, with the source document cited. Everything is in the GitHub repo agentic-devops What RAG Is — Without the Hype RAG stands for Retrieval-Augmented Generation. Instead of asking an LLM a question and hoping its training data contains the answ

Vercel Labs recently open-sourced zero-native, a cross-platform framework for native desktop applications. Zero-native bypasses Electron runtime in favor or native OS WebViews and claims to achieve smaller, more efficient native apps with minimal overhead. Zero-native is written in Zig, thus directly interoperates with native C libraries, and features fast incremental compilation times. By Bruno Couriol

MVCC, WAL, vacuum, and replication slots under sustained delete load - and how to delete billions of rows without your database noticing Most "how to delete a lot of rows" articles stop at "batch it and delete children before parents." That advice is correct, it's table stakes, and everyone already knows it. This article is about everything after that - the parts that actually decide whether your cleanup runs quietly in the background for a week or pages you at 3 a.m. with a full disk and a replica that's six hours behind. The thesis: at scale, your DELETE statement is the easy part. The adversaries are the subsystems a delete feeds - MVCC tuple versioning, the write-ahead log, autovacuum, and the replication machinery. Bulk deletion is really an exercise in flow control across those subsystems . Get the SQL right and the systems wrong, and you'll still take production down. We'll assume PostgreSQL (the internals are PG-specific), a live OLTP primary with at least one physical replica and one or more logical/CDC consumers, and a target of hundreds of millions to billions of rows across many related tables. The one paragraph of "basics," so we can move on: delete in dependency order (referencing rows before referenced rows); collect parent keys once; never rely on ON DELETE CASCADE for huge deletes because you can't throttle a cascade. Done. Now the real material. 1. What a DELETE actually costs A delete is not "remove a row." Under MVCC it's "mark a row version dead and write that fact everywhere." For each deleted tuple, PostgreSQL: Sets xmax on the heap tuple to your transaction id. The row is still physically present; it becomes a dead tuple once your transaction commits and no snapshot can still see it. Writes a WAL record for the heap change. If this is the first modification of that page since the last checkpoint, it also writes a full-page image (FPI) - potentially 8 KB of WAL for a single-row change. Touches every index. Index entries aren't removed at delet

Every backend engineer has seen this happen, you build an application on a relational database like MySQL, handling thousands of concurrent transactions effortlessly. Then, the business asks for a real time analytics dashboard. But when you run an aggregation query over historical data, suddenly the database that effortlessly managed live traffic starts thrashing, evicting your working set, and dragging application performance down. This isn't a tuning problem, a missing index, or a badly written query. It’s a fundamental architectural collision. OLTP (Online Transaction Processing) OLTP encompasses nearly every concurrent digital interaction triggered across a distributed system. A user downloading a PDF, a microservice firing an automatic maintenance log, a comment on a social feed these are all transactions. Data engineers rely on OLTP systems (like MySQL or PostgreSQL) to capture these concurrent streams of interactions for creating , updating and deleting records. The Tree Based In-Place Engine To reliably capture massive volumes of transactions without corrupting data or locking up the application, OLTP systems rely on a highly optimized, row oriented architecture built around the B+ Tree. Because they must provide immediate, atomic updates to existing records, transactional databases manage state through a strict sequence of physical tree traversal and in-memory page mutation: The B+ Tree Indexing: When a transaction reads or updates id: 1, the engine traverses a B+ Tree from the root, through the branch nodes, directly to the specific physical leaf node holding that row. This O(\log n) traversal guarantees a fast, isolated point-lookup. It ensures the application always hits the single version of the row without scanning irrelevant data. The Buffer Pool & In-Place Updates: OLTP systems perform in place updates. The database pulls the exact page containing id: 1 from the physical disk into memory (the Buffer Pool). The specific row is mutated directly in RAM