#jwt — Public Fediverse posts

Does an encrypted as well as signed web token make sense? Clearly, why are these almost none? Clearly, signed are unchangeable but also via HTTPS they say a lot about the user what (depending on the). It is not about security but about confidentiality.

#thoughts #token #websecurity #mitm #itsecurity #web #online #webdev #https #opinion #jwt #paseto

Does an encrypted as well as signed web token make sense? Clearly, why are these almost none? Clearly, signed are unchangeable but also via HTTPS they say a lot about the user what (depending on the). It is not about security but about confidentiality.

#thoughts #token #websecurity #mitm #itsecurity #web #online #webdev #https #opinion #jwt #paseto

Black Hole Star = #JWT Red dots. Start #writing!

https://news.mit.edu/2026/astronomers-discover-brand-new-type-astrophysical-object-black-hole-star-0812

#BoostingIsSharing

#sf #scienceFiction #astrophysics #science #Writingcommunity #writersOfMastodon #blackholes

Black Hole Star = #JWT Red dots. Start #writing!

https://news.mit.edu/2026/astronomers-discover-brand-new-type-astrophysical-object-black-hole-star-0812

#BoostingIsSharing

#sf #scienceFiction #astrophysics #science #Writingcommunity #writersOfMastodon #blackholes

Rapid7 published full details and a PoC for CVE-2026-55040, a critical SharePoint authentication bypass that forges JWT tokens. Patch now.

#CVE202655040 #SharePoint #AuthenticationBypass #JWT #Rapid7 #PoC #Cybersecurity

https://securityonline.info/sharepoint-auth-bypass-cve-2026-55040/?utm_source=mastodon&utm_medium=jetpack_social

Rapid7 published full details and a PoC for CVE-2026-55040, a critical SharePoint authentication bypass that forges JWT tokens. Patch now.

#CVE202655040 #SharePoint #AuthenticationBypass #JWT #Rapid7 #PoC #Cybersecurity

https://securityonline.info/sharepoint-auth-bypass-cve-2026-55040/?utm_source=mastodon&utm_medium=jetpack_social

WSO2 Universal Gateway v4.5.0 & 4.6.0 affected by CRITICAL CVE-2026-5430 (CVSS 10.0). Improper JWT validation enables account takeover. No patch yet — apply compensating controls. https://radar.offseq.com/threat/cve-2026-5430-cwe-347-improper-validation-of-certificate-with-host-mismatch-in-wso2-wso2-universal-f913080655427dd8 #OffSeq #WSO2 #JWT #Vulnerability

Архитектура под замену: чему нас научил переход с OIDC на Kerberos

Нужно было заменить OIDC на Kerberos при внедрении платформы в облако заказчика. Простая замена не сработала: пользовательский контекст и межсервисные вызовы оказались плотно завязаны на OIDC. В статье рассказываю, как мы искали компромиссное решение, пересматривали границы между бизнес-логикой и инфраструктурой и довели проект до приемки без полной переработки платформы. Что пошло не так

https://habr.com/ru/companies/cloud_ru/articles/1065624/

#Kerberos #OIDC #JWT #IAM #микросервисы #аутентификация #токены #архитектура #облако

i dont get much of a reason to use #JWT s at least on projects meant to be self hosted, like api tokens that are stored in db achives everything without complex renewal and revocation logic and when you need to support old protocols or simpler clients you need to implement app password kind of system and come back to needing to store hashes of tokens anyway

OAuth‑сервер, который не хранит пользователей

Наш OAuth-сервер не хранит профиль пользователя и практически ничего о нём не знает. Он знает только IdP-личность («вы вошли через Google под таким-то аккаунтом»), а внутренний id , роль и имя живут в продуктовом бэкенде. Сшивают их два кастомных гранта: первый вкладывает в токен подписанный профиль, не зная его содержимого, а второй разменивает основной токен на веер узких производных, по одному на аудиторию. Кроме того, в статье объясняется, почему FedCM может создавать больше проблем, чем решать, и как эти проблемы обходятся с помощью Service Worker. Если вам интересно, как всё это работает в SaaS, прошу под кат.

https://habr.com/ru/articles/1063816/

#OAuth_20 #OpenID_Connect #JWT #FedCM #Service_Worker #token_exchange #PKCE #JWKS

Test JWT clock skew tolerance by manually shifting system time with timedatectl on Ubuntu 22.04, Debian 12, or RHEL 9. Validate tokens with known iat, nbf, exp timestamps against tolerance windows (e.g., 30s) and log discrepancies. #jwt #clock-skew #testing

https://www.valtersit.com/vault/test-jwt-for-clock-skew-tolerance-with-manual-time-adjustmen-230318/

CVE-2026-45363: CRITICAL improper authentication in ruby-jwt (<2.10.3, <3.2.0) lets attackers forge JWTs with empty HMAC keys. Immediate upgrade required to 2.10.3/3.2.0. Impact: auth bypass, high integrity risk. https://radar.offseq.com/threat/cve-2026-45363-cwe-287-improper-authentication-in--588cdab561cd07ab #OffSeq #CVE202645363 #ruby #jwt #infosec

Жизненный цикл API-токенов: От генерации до компрометации

JWT не защищает от кражи. Подпись гарантирует, что токен не подделан, но не мешает злоумышленнику использовать его, если тот оказался в чужих руках. В 2026 году утечки токенов через XSS, логи и CDN-атаки входят в тройку самых частых векторов компрометации, а медианное время обнаружения такой кражи составляет 43 дня. В этой статье разбираем полный жизненный цикл токена — от выдачи до отзыва. Сравниваем хранение в localStorage , HttpOnly -куках и BFF-прослойке. Объясняем, почему Refresh Token Rotation — это must-have, а не опция, и как детектировать кражу еще до того, как злоумышленник успеет навредить.

https://habr.com/ru/articles/1058318/

#jwt #jwt_token #jwt_auth #jwt_authentication #appsec #security #frontend #frontend #frontend_разработка #frontendразработка

Blacklist revoked JWTs with Redis using the token's jti claim as key and expiry matching remaining lifetime. Check blacklist on every API request before signature verification. Stateless token handling, consistent revocation. #jwt #snippet #revocation #ValtersIT

https://www.valtersit.com/vault/blacklist-compromised-jwts-with-redis-and-consistency-checks-9a065b/

Тёмная сторона Go: разбор живых уязвимостей с продакшена и инструменты против них

Тёмная сторона есть, наверное, у каждого языка. У Go она выражена слабее: язык структурный, понятный, неплохо защищён по умолчанию. Но это не индульгенция. Разберём, что такое уязвимость и откуда она берётся, пройдёмся по типичным классам проблем на живых примерах с продакшена, посмотрим на инструменты локального анализа кода и обсудим, как выстроить продуктивные отношения с командой безопасности.

https://habr.com/ru/companies/oleg-bunin/articles/1048122/

#Go #golang #информационная_безопасность #безопасная_разработка #HTTP_request_smuggling #JWT #логирование #govulncheck #DevSecOps #OWASP

Software engineer Mukesh Murugan writes this recent first in-depth article in a three-part series covering roles, claims, and policies for JWT-based authentication, and authorization in ASP.NET Core web apis using C#.

His blog (https://codewithmukesh.com/) is high signal-to-noise ratio, on par with the likes of C# Corner. Highly recommend.

"Role-Based Authorization in ASP.NET Core - A .NET 10 Guide"

https://codewithmukesh.com/blog/role-based-authorization-in-aspnet-core/

#dotnet #csharp #rbac #webapi #jwt #programming

Software engineer Mukesh Murugan writes this recent first in-depth article in a three-part series covering roles, claims, and policies for JWT-based authentication, and authorization in ASP.NET Core web apis using C#.

His blog (https://codewithmukesh.com/) is high signal-to-noise ratio, on par with the likes of C# Corner. Highly recommend.

"Role-Based Authorization in ASP.NET Core - A .NET 10 Guide"

https://codewithmukesh.com/blog/role-based-authorization-in-aspnet-core/

#dotnet #csharp #rbac #webapi #jwt #programming

🔑 JWT is a scam and your app doesn't need it

｢ The token is valid until it expires, full stop. The only way to invalidate it is to store the jti server-side in a revocation list and check that list on every single request. Which is a database lookup. Which is the thing JWT was supposed to let you skip. Congratulations, you've reinvented sessions, badly ｣

https://www.dusanmalusev.dev/blog/jwt-is-a-scam-and-your-app-doesnt-need-it

#jwt #authentication

🔑 JWT is a scam and your app doesn't need it

｢ The token is valid until it expires, full stop. The only way to invalidate it is to store the jti server-side in a revocation list and check that list on every single request. Which is a database lookup. Which is the thing JWT was supposed to let you skip. Congratulations, you've reinvented sessions, badly ｣

https://www.dusanmalusev.dev/blog/jwt-is-a-scam-and-your-app-doesnt-need-it

#jwt #authentication

Stop Using JWTs

https://gist.github.com/samsch/0d1f3d3b4745d778f78b230cf6061452

#HackerNews #StopUsingJWTs #JWT #Security #WebDevelopment #Cybersecurity #TechDebate

Stop Using JWTs

https://gist.github.com/samsch/0d1f3d3b4745d778f78b230cf6061452

#HackerNews #StopUsingJWTs #JWT #Security #WebDevelopment #Cybersecurity #TechDebate

ID, token, UUID и slug: в чём разница и почему их нельзя мешать

В каждом бэкенде рано или поздно рядом появляются id , UUID , slug , token и request_id . Все они выглядят как строки, но отвечают за разные вещи. Когда это забывают, UUID становится защитой, slug — вечной айдишкой, а token — просто ещё одним идентификатором.

https://habr.com/ru/articles/1048324/

#backend #api #архитектура #uuid #jwt #token #slug #session_id #request_id #идентификаторы

La debilidad de #JWT está en conocer la clave secreta 🔑

https://leviathan.3xi.club/jwt
ever curious about the size impact of algorithm choice? esp when quantum-resistant algos join the mix? explore EdDSA, ES256, ML-DSA, SLH-DSA-SHAKE, SLH-DSA-SHAKE, MLDSA44-SLHDSA classical and post-quantum JWT signatures in this interactive demo
#crypto #cryptography #jwt #tokens

Пишем движок для блога на Rust

Недавно я решил завести собственный блог. Сначала посмотрел в сторону SSG, но они показались мне не слишком удобными для того сценария, который я хотел получить. Затем попробовал несколько CMS, однако быстро упёрся в другую проблему: мой сервер оказался слишком слабым для большинства современных решений. В итоге ни одно из готовых решений так и не смогло закрыть все мои требования одновременно. Так и появилась идея сделать небольшую файловую CMS на Rust, которая не требует базы данных, не потребляет много памяти и при этом остаётся достаточно гибкой для повседневного использования. Со временем идея небольшого блогового движка разрослась в полноценную CMS с SSR, виртуальной файловой системой, поддержкой локализации, визуальным редактором статей и горячей перезагрузкой контента. В этой статье я постараюсь показать, как всё это устроено изнутри.

https://habr.com/ru/articles/1041324/

#rust #tokio #ssr #vfs #typescript #rxjs #jwt #fluent #kdl

JWT: Just a #Wobbly Trick! 🤡 It's less about authenticating and more about complicating. But hey, if you love slow, leaky, unmanageable code with a side of developer confusion, #JWT is truly your jam! 🚀🔧
https://www.dusanmalusev.dev/blog/jwt-is-a-scam-and-your-app-doesnt-need-it #Trick #DeveloperConfusion #CodeQuality #SoftwareDevelopment #HackerNews #ngated

JWT: Just a #Wobbly Trick! 🤡 It's less about authenticating and more about complicating. But hey, if you love slow, leaky, unmanageable code with a side of developer confusion, #JWT is truly your jam! 🚀🔧
https://www.dusanmalusev.dev/blog/jwt-is-a-scam-and-your-app-doesnt-need-it #Trick #DeveloperConfusion #CodeQuality #SoftwareDevelopment #HackerNews #ngated

JWT is a scam and your app doesn't need it

https://www.dusanmalusev.dev/blog/jwt-is-a-scam-and-your-app-doesnt-need-it

#HackerNews #JWT #scam #app #security #developer #insights

JWT is a scam and your app doesn't need it

https://www.dusanmalusev.dev/blog/jwt-is-a-scam-and-your-app-doesnt-need-it

#HackerNews #JWT #scam #app #security #developer #insights

⚠️ CRITICAL: CVE-2026-44699 in libjwt (3.0.0 – 3.3.2) lets attackers forge JWTs by exploiting HMAC verification with a zero-length key if RSA JWKs lack the alg parameter. Patch to 3.3.3 or confirm vendor remediation! https://radar.offseq.com/threat/cve-2026-44699-cwe-327-use-of-a-broken-or-risky-cr-78313648 #OffSeq #JWT #Infosec

LHS 3844 b: Das James-Webb-Weltraumteleskop hat einen Exoplaneten analysiert https://www.mpg.de/26331922/astronomen-untersuchen-die-oberflaechenbeschaffenheit-einer-nahen-supererde -- Na dann gute Nacht Internet.

#jwt #LHS3844b

Логин через Telegram по-новому: разбираем OIDC-флоу oauth.telegram.org и собираем его на Python

Telegram теперь полноценный OpenID-провайдер: oauth.telegram.org, JWKS, JWT, claims. Туториалы на GitHub при этом массово показывают старый виджет с HMAC от bot-token и /setdomain в BotFather. Я разобрался с новым флоу и собрал PoC на Python — рассказываю, как устроен обмен между фронтом, Telegram и бэком, чем Login library через telegram-login.js отличается от manual OIDC code flow с PKCE, что настраивать в BotFather (спойлер: не в чате, а в его mini-app), как протестировать локально через ngrok, и какая проверка id_token нужна вместо ручного HMAC.

https://habr.com/ru/articles/1033632/

#telegram #telegram_login #openid_connect #oidc #oauth2 #jwt #jwks #pyjwt #python #authentication

Я устал от Duolingo и написал себе AI-репетитора. Go, Clean Architecture, 4 LLM-модели — и вот что из этого вышло

Мой рабочий день - это код. Вечером я хочу разговаривать с кем-то по-английски, а не нажимать на пингвинчиков. Duolingo учит меня заказывать яблоки, ChatGPT-чат отлично объясняет грамматику - но не помнит, что я разбирал Present Perfect в среду и опять путаю его с Past Simple в пятницу. Я хотел простую штуку: написать модели «давай сегодня про багтрекеры»; получить чат на 15 минут; а в конце - три новых слова , которые она же мне и подобрала по уровню B1. Чтобы завтра эти слова всплыли в упражнениях. Через месяц получился Lexis : Go · Clean Architecture · 4 LLM-модели · SSE-стриминг · JWT rotation + reuse detection MIT, открытый репозиторий. В статье - три инженерных якоря , которыми я доволен: Pluggable AI-провайдеры через интерфейс из трёх методов. SSE вместо WebSocket для стриминга AI-ответов. JWT rotation + reuse detection как production-ready стандарт, а не «потом перепишем». Плюс честный список того, что ещё не готово. Показать архитектуру

https://habr.com/ru/articles/1030928/

#Go #Clean_Architecture #JWT #AI #LLM #petproject #изучение_английского #SSE #modular_monolith #refresh_tokens

Почему ваша LLM-платформа — следующая цель: аудит безопасности AI-сервиса изнутри

Мы искали уязвимости в RAG-платформе с десятками тысяч пользователей — а нашли доступ ко всей инфраструктуре и API-ключам с бюджетом в сотни тысяч долларов. Две недели мы строили сложные цепочки: SSRF через LangChain, инъекции в промпты, HTTP smuggling, CVE в десериализации. Ни одна не дала результата. А потом мы сделали один curl к открытому порту — и получили все ключи за 5 минут. Эта статья — не гайд по взлому . Это разбор того, почему LLM-инфраструктура создаёт принципиально новые риски, какие ошибки мы раз за разом видим в AI-стартапах, и на что стоит обратить внимание, если вы строите что-то похожее.

https://habr.com/ru/articles/1029822/

#пентест #LLM #SSRF #JWT #Docker #LangChain #AI_Security #аудит_безопасности #RAG #APIключи

🚨 CRITICAL: CVE-2026-6911 in AWS Ops Wheel — missing JWT signature checks allow unauth access & admin control over all tenants. Patch by redeploying from the updated repo! Details: https://radar.offseq.com/threat/cve-2026-6911-cwe-347-improper-verification-of-cry-0f0da004 #OffSeq #AWS #Vuln #JWT

Contentful is hiring Security Engineer (Application Security)

🔧 #golang #javascript #python #aws #jwt #kubernetes #terraform #securityengineer
🌎 Atlanta, Georgia
⏰ Full-time
🏢 Contentful

Job details https://jobsfordevelopers.com/jobs/security-engineer-application-security-at-contentful-com-apr-6-2026-baadb6?utm_source=mastodon.world&utm_medium=social&utm_campaign=posting
#jobalert #jobsearch #hiring

I'm planning to do a lightning ⚡ talk tonight - To JWT or not to JWT - why JSON Web Tokens might not be the one size that fits all (in my opinion)

vote for me in the #DjangoCon Chat:
https://chat.djangocon.eu/#narrow/channel/16-Lightning-talks/topic/Thursday/near/913

#selfpromotion #django #jwt #security

【セキュリティ】JWTの署名を信頼していたら `alg: none` で管理者権限を奪取された話
https://qiita.com/fe1ix/items/0b1ec01a6a41de94efa5?utm_campaign=popular_items&utm_medium=feed&utm_source=popular_items

#qiita #Security #Web #初心者 #JWT #新人プログラマ応援

Novo post no blog: JWT: três erros que todo mundo comete na primeira implementação

Você copia o exemplo do PyJWT, funciona, e acha que está pronto. Mas a implementação padrão tem três problemas sérios: aceitar o algoritmo que o token declara, ausência de revogação, e segredo fraco.

Para cada erro: o mecanismo, como explorar, e como corrigir.

https://riverfount.dev.br/posts/jwt_tres_erros/

#python #segurança #jwt #autenticação

Lucee in a Box: The Ultimate Guide to Containerized Dev Servers

2,726 words, 14 minutes read time.

The Modern ColdFusion Workspace: Transitioning to Lucee in a Box

The shift from traditional, monolithic server installations to containerized environments has fundamentally altered how we perceive modern development within the Lucee ecosystem. For years, the standard approach involved installing a heavy application server directly onto a local machine, often leading to a “polluted” operating system where various versions of Java and Lucee competed for resources and environment variables. By adopting a “Lucee in a Box” methodology, we decouple the application logic from the underlying hardware, allowing for a portable, reproducible, and lightweight development stack. This transition is not merely about convenience; it is a strategic move toward parity with production environments where high availability and rapid scaling are the norms. In this architecture, we utilize Docker to encapsulate the Lucee engine, the web server, and the necessary configuration files into a single unit that can be spun up or destroyed in seconds, ensuring that every member of a development team is working within an identical, script-driven environment.

However, the true complexity of this setup emerges when we move beyond simple “Hello World” examples and begin integrating with the existing corporate infrastructure. In my own workflow, I rely heavily on a network of internal web services that act as the primary conduit for data residing in our production databases. These services are vital because they provide a sanitized, governed layer of abstraction over raw SQL queries, ensuring that sensitive data is handled according to internal compliance standards. When we containerize Lucee, we aren’t just running a script; we are placing a small, isolated node into a complex network. The challenge then becomes ensuring this isolated container can “see” and communicate with those internal services as if it were a native part of the network, all while maintaining the security boundaries that containerization is designed to provide.

The Data Silo Crisis: Overcoming Networked Service Isolation

One of the most significant hurdles in modernizing a CFML stack is the inherent isolation of the Docker bridge network, which often creates what I call a “Data Silo” during local development. When a developer attempts to call an internal web service—perhaps a REST API that fetches real-time production metrics or user permissions—from within a container, the request often hits a wall because the container’s internal DNS does not naturally resolve local intranet addresses. This creates a frustrating disconnect where the application works perfectly in the legacy local install but fails within the containerized environment. This disconnect is more than a minor annoyance; it leads to significant delays in the development lifecycle as engineers struggle to pipe in the data necessary for testing complex business logic. Without a seamless connection to these internal services, the “Lucee in a Box” becomes an empty vessel, incapable of performing the data-intensive tasks required in a modern enterprise setting.

To resolve this, we must look at how the container perceives the outside world and how the host machine facilitates that visibility. In many corporate environments, production data is guarded behind strict firewall rules and SSL requirements that expect requests to originate from known entities. When I utilize internal web services to provide data from a production database, the Lucee container must be configured to pass through the host’s network or be explicitly granted access to the internal DNS suffixes. Failure to address this at the architectural level results in “unreachable host” errors or SSL handshake failures that can derail a project for days. By understanding that the container is a guest on your network, we can begin to implement the routing and trust certificates necessary to turn that siloed container into a fully integrated node capable of consuming live data streams securely and efficiently through modern CFScript syntax.

The Blueprint: Implementing Lucee and MariaDB via Docker Compose

To move from theory to implementation, we must define the orchestration layer that brings our environment to life. The docker-compose.yml file is the definitive source of truth for the development stack, eliminating the “it works on my machine” excuse by codifying the server version, database configuration, and network paths. In the professional workflow I advocate, this file sits at the root of your project. It defines a lucee service using the official Lucee image—optimized for performance—and a mariadb service to handle local data persistence. Crucially, we use volumes to map your local www folder directly into the container’s web root. This means that as you write your CFScript in your preferred IDE on your host machine, the changes are reflected instantly inside the container without requiring a rebuild or a manual file transfer.

The following configuration provides a professional-grade starting point. It establishes a dedicated network for our services and ensures that Lucee has the environment variables necessary to eventually automate its datasource connections. By mounting the ./www directory, we ensure our code remains on our host machine where it can be version-controlled, while the ./db_data volume ensures our MariaDB data persists even if the container is destroyed and recreated.

version: '3.8'

services:
  # The Database Engine
  mariadb:
    image: mariadb:10.6
    container_name: lucee_db
    restart: always
    environment:
      MYSQL_ROOT_PASSWORD: root_password
      MYSQL_DATABASE: dev_db
      MYSQL_USER: dev_user
      MYSQL_PASSWORD: dev_password
    volumes:
      - ./db_data:/var/lib/mysql
    networks:
      - dev_network

  # The Lucee Application Server
  lucee:
    image: lucee/lucee:5.3
    container_name: lucee_app
    restart: always
    ports:
      - "8080:8888"
    environment:
      # Injecting DB credentials for CFConfig or Application.cfc
      - DB_HOST=mariadb
      - DB_NAME=dev_db
      - DB_USER=dev_user
      - DB_PASSWORD=dev_password
      - LUCEE_ADMIN_PASSWORD=server_admin_pass
    volumes:
      - ./www:/var/www
      - ./config:/opt/lucee/web
    depends_on:
      - mariadb
    networks:
      - dev_network

networks:
  dev_network:
    driver: bridge

Deployment Strategy: Running Your New Containerized Stack

Once the docker-compose.yml file is in place, initializing the environment is a matter of a single terminal command. By executing docker-compose up -d from the root of your project directory, the Docker engine pulls the specified images, creates the isolated virtual network, and establishes the volume mounts. This process ensures that your MariaDB instance is ready to receive connections before the Lucee server fully initializes. For developers who rely on internal web services, this is where the containerized approach proves its worth. Because Lucee is running in an isolated network but can be configured to have access to the host’s bridge or external DNS, it can safely consume external APIs while maintaining a clean, local database for session state or cached production data. This setup provides the exact same architectural “feel” as a high-traffic production cluster, but contained entirely within your local hardware.

The beauty of this system lies in its maintenance-free nature and the elimination of the “dependency hell” that often plagues legacy ColdFusion developers. If you need to test your CFScript against a different version of Lucee or a newer patch of MariaDB, you simply update the version tag in the YAML file and run the command again. There is no need to uninstall software, clear registry keys, or worry about Java version conflicts on your host machine. This modularity is why I utilize internal web services to provide data from production into this local box; the container acts as a secure, high-speed proxy. You can pull the data you need via an internal API call, store it in the MariaDB container, and work in an isolated state without ever risking the integrity of the actual production database.

Root Cause: Why Standard Containers Fail at Internal Service Integration

The primary reason most off-the-shelf Lucee container configurations fail when attempting to consume internal web services is a fundamental lack of trust—specifically, the absence of internal SSL certificates within the Java KeyStore. When I use web services hosted within my network to provide data from a production database, those services are almost always secured via an internal Certificate Authority (CA) that is not recognized by the default OpenJDK installation inside the Lucee container. This results in the dreaded “PKIX path building failed” error the moment a cfhttp call is initiated via CFScript to an internal endpoint. To solve this, the Dockerfile must be modified to perform a “copy and import” operation during the image build phase, where the internal CA certificate is added to the Java security folder and registered using the keytool utility. This ensures that the underlying Java Virtual Machine (JVM) trusts the internal network’s identity, allowing for encrypted, secure data transmission from the production-proxy services to the local development environment.

Beyond the cryptographic hurdles, there is the issue of routing and “Host-to-Container” communication that often stymies developers new to the Docker ecosystem. In a standard Docker setup, the container is wrapped in a layer of Network Address Translation (NAT) that makes it difficult to reach services sitting on the developer’s physical host or the wider corporate VPN. To bridge this gap, we often utilize the extra_hosts parameter within our docker-compose configuration, which effectively injects entries into the container’s /etc/hosts file. This allows us to map a friendly internal domain name, like services.internal.corp, directly to the IP address of the host machine or the VPN gateway. By explicitly defining these routes, we bypass the limitations of Docker’s isolated bridge and enable the Lucee engine to reach out to the web services that house our production data. This architectural “handshake” between the containerized Lucee instance and the physical network is the secret sauce that transforms a basic dev box into a high-fidelity replica of the production ecosystem.

Deep Dive: Consuming Internal Web Services via CFScript

With the network and security infrastructure in place, we can finally focus on the implementation layer: the CFScript that handles the data exchange. In a modern Lucee in a Box setup, I favor a service-oriented architecture where a dedicated DataService.cfc handles all interactions with the internal network. Using the http service in CFScript, we can construct requests that include the necessary authentication headers, such as JWT tokens or API keys, required by the internal production data services. The beauty of this approach is that the CFScript remains agnostic of the container’s physical location; as long as the Docker networking layer is correctly mapping the service URL to the internal network, the cfhttp call proceeds as if it were running on a native server. This allows us to maintain a clean, readable codebase that utilizes the latest CFScript features, such as cfhttp(url=targetURL, method="GET", result="local.apiResponse"), while the heavy lifting of network routing is handled by the Docker daemon.

The real power of this integration is realized when we use these internal web services to populate our local MariaDB instance with a “snapshot” of production-like data. Rather than dealing with massive, cumbersome database dumps that can compromise data privacy, we can write an initialization script in CFScript that queries the internal web services for the specific datasets required for a given task. This script can then parse the returned JSON and perform a series of queryExecute() commands to populate the local MariaDB container. This “just-in-time” data strategy ensures that the developer is always working with relevant, fresh data without the security risks associated with a direct connection to the production database. By leveraging the containerized Lucee instance as a smart bridge between internal network services and local storage, we create a development environment that is not only isolated and secure but also incredibly data-rich and performant.

Environment Variable Injection: The CFConfig and CommandBox Synergy

To achieve a truly “hands-off” configuration within a Lucee in a Box environment, we must move away from the manual web-based administrator and toward a purely scripted setup. This is where the combination of CommandBox and the CFConfig module becomes indispensable. By using a .cfconfig.json file or environment variables prefixed with LUCEE_, we can define our MariaDB datasource connections, internal web service endpoints, and mail server settings without ever clicking a button in the Lucee UI. In a professional workflow, this means the docker-compose.yml file serves as the master controller, injecting credentials and network paths directly into the Lucee engine at runtime. For instance, by setting LUCEE_DATASOURCE_MYDB as an environment variable, the containerized engine automatically constructs the connection to the MariaDB container, ensuring that our CFScript-based queryExecute() calls have a reliable target the moment the server is healthy.

This approach is particularly powerful when dealing with the internal web services that provide our production data. Since these services often require specific API keys or internal proxy settings, we can store these sensitive values in an .env file that is excluded from our Git repository. When the container starts, these values are mapped into the Lucee process, allowing our CFScript logic to access them via system.getEnv(). This ensures that our local development environment remains a mirror of our production logic while maintaining a strict separation of concerns between the application code and the infrastructure-specific secrets. By automating the configuration layer, we eliminate the risk of manual setup errors and ensure that every developer on the team can spin up a fully functional, networked-aware Lucee instance in a single command.

Advanced Networking: Bridged Access to Production-Proxy Services

The final piece of the Lucee in a Box puzzle involves fine-tuning the Docker network to handle the high-latency or high-security requirements of internal web services. When our CFScript makes a request to a service that pulls from a production database, we are often traversing multiple layers of internal routing, including VPNs and load balancers. To optimize this, we can configure our Docker bridge network to use specific MTU (Maximum Transmission Unit) settings that match our corporate network’s infrastructure, preventing packet fragmentation that can lead to mysterious request timeouts. Furthermore, by utilizing Docker’s aliases within the network configuration, we can simulate the production URL structure locally. This means our CFScript can call https://api.internal.production/ both in the dev container and the live environment, with Docker handling the redirection to the appropriate internal service endpoint based on the environment context.

Beyond simple connectivity, we must also consider the performance of these data-heavy web service calls. In a containerized environment, I often implement a caching layer within Lucee that stores the JSON payloads returned from our internal services into the local MariaDB instance or a RAM-based cache. By using CFScript’s cachePut() and cacheGet() functions, we can significantly reduce the load on our internal network and the production database proxy. This “lazy-loading” strategy allows us to develop complex features with the speed of local data access while still maintaining the accuracy of production-sourced information. This architectural decision—balancing live service integration with local persistence—represents the pinnacle of the Lucee in a Box philosophy, providing a development experience that is as fast as it is faithful to the real-world environment.

Conclusion: The Future of Scalable CFML Development

Adopting a “Lucee in a Box” strategy is more than just a trend in containerization; it is a fundamental shift toward professional-grade, reproducible engineering. By strictly defining our environment through docker-compose.yml, automating our security through SSL injection in the Dockerfile, and utilizing CFScript to bridge the gap between internal web services and local MariaDB storage, we create a stack that is resilient to “configuration drift.” This setup allows us to treat our development servers as ephemeral, disposable assets that can be rebuilt at a moment’s notice to match evolving production requirements. As the Lucee ecosystem continues to mature, the ability to orchestrate these complex data flows within a containerized boundary will remain the hallmark of a high-performing development team, ensuring that we spend less time debugging infrastructure and more time writing the logic that drives our applications forward.

Call to Action

If this post sparked your creativity, don’t just scroll past. Join the community of makers and tinkerers—people turning ideas into reality with 3D printing. Subscribe for more 3D printing guides and projects, drop a comment sharing what you’re printing, or reach out and tell me about your latest project. Let’s build together.

D. Bryan King

Sources

• Lucee Official Documentation: Lucee Docker Images
• Ortus Solutions: CommandBox Docker Documentation
• NIST SP 800-190: Application Container Security Guide
• Docker Documentation: Bridge Network Driver and Architecture
• IANA: Service Name and Transport Protocol Port Number Registry
• MITRE CWE-601: URL Redirection to Untrusted Site (Internal Service Security)
• Docker Engine: Dockerfile Reference and Best Practices
• CFConfig: Environment Variable Integration for Lucee
• CIS Benchmarks: Docker Community Edition Benchmark
• IETF RFC 7519: JSON Web Token (JWT) for Internal Service Auth
• Wireshark User Guide: Analyzing Internal Network Traffic
• Oracle Java SE Security: Standard Algorithm Names (TLS Support for Lucee)

Disclaimer:

The views and opinions expressed in this post are solely those of the author. The information provided is based on personal research, experience, and understanding of the subject matter at the time of writing. Readers should consult relevant experts or authorities for specific guidance related to their unique situations.

Related Posts

Rate this:

Ever pasted a JWT into jwt.io? That token — with your user IDs, roles, and claims — just went to someone else's server.

Our JWT Decoder runs entirely in your browser. Paste, decode, done. Nothing leaves your device.

https://dannycranmer.github.io/devtoolbox/jwt-decoder.html

Full tutorial: https://dev.to/danny_cranmer_9e9e9d7178d/how-to-debug-jwt-tokens-without-sending-them-to-a-server-5826

#privacy #security #webdev #jwt #opensource #devtools

Ah, the timeless pursuit of "dark mode" 🤡—because who needs actual functionality when you can squint in style? And let's not forget the epic battle of #JWT vs. #sessions, a showdown sure to keep absolutely no one on the edge of their seat. Just #npm install and let the over-engineering commence! 🎉
https://rjcorwin.github.io/cook/ #darkmode #overengineering #HackerNews #ngated