Backend et Fullstack

Backend & Fullstack — 8 septembre 2026

Lo nuevo hoy

Today's highlights

Points forts du jour

Click en cualquiera para ir al detalle

Click any item to jump to the full section

Cliquez un élément pour aller à la section complète

🔥

Top Stories

Destacado

Netflix tira su autoscaler propio y adopta el de Apache Flink: 1,1 millones menos por año

El autoscaler propietario de Netflix escalaba a nivel de cluster: aplicaba la misma decisión a todos los operators de un job. Eso funciona hasta que tenés pipelines con branches, joins y terabytes de estado, donde cada componente tiene necesidades de procesamiento distintas. El autoscaler de Apache Flink analiza vertex por vertex, calculando el paralelismo requerido de cada uno a partir de throughput y busy time.

Los números: un equipo bajó 58% el gasto anualizado de cómputo Flink, ~1,1 millones de dólares al año. Netflix corre más de 30.000 streaming jobs en varias regiones de AWS.

Dos detalles de ingeniería que valen más que el titular. Primero, integraron el autoscaler con su control plane usando Spring Boot y workflows de Temporal para aislar las decisiones de escalado por job, y tuvieron que mejorar la recolección de métricas del JobManager para soportar jobs de hasta 3.000 subtasks, con filtrado de métricas del lado del server. Segundo, usan un target de utilización de 0.45 en vez del 0.7 por defecto de la comunidad, justamente para no reescalar agresivamente aplicaciones stateful grandes — mover terabytes de estado cuesta más que la capacidad que ahorrás.

07 Sep 2026
infoq.com →
Destacado

Airbnb mueve la autenticación al server y borra el 60% del código

La idea es simple de enunciar y difícil de aceptar: en vez de que el cliente decida qué método de autenticación mostrar, el server toma la decisión y el cliente renderiza la pantalla que el server devuelve. Eso les permite preferir OTP por WhatsApp en Brasil o proveedores de identidad locales en Corea del Sur sin shipear código nuevo de cliente.

El 60% menos de código sale de la centralización: web, iOS y Android renderizan pantallas estandarizadas definidas por el server en vez de embeber la lógica, y ninguno duplica la selección de challenge. Además los schemas del server generan las definiciones de tipos del cliente. Resultado: 100 KB menos de bundle web y ~11% de ahorro en costos de OTP.

Los números de producto son fuertes: la experimentación pasó de semanas a días (20+ experimentos en tres meses), 2,6% de mejora en éxito de autenticación y 27% menos cuentas duplicadas.

Ahora, los tradeoffs son reales y no los escondas: más complejidad del lado del server manejando policy engines y lógica regional, y el cliente queda dependiendo de la disponibilidad del server en cada paso del login. Eso último no es menor — te convertiste en un single point of failure para la puerta de entrada de tu producto. Si vas por este camino, el diseño de degradación es parte del diseño, no un extra.

04 Sep 2026
infoq.com →
Beta

KYAML: un subset estricto de YAML para dejar de adivinar tipos en Kubernetes

KYAML es un subset estricto de YAML: objetos con {}, arrays con [], y strings con comillas dobles obligatorias. Conserva lo útil de YAML —los comentarios— y le saca la ambigüedad: estructura por indentación que se rompe sola, y valores sin comillas que se interpretan como otro tipo. Ese problema se multiplica cuando los manifests los genera un sistema de templating como Helm, que es donde más duele.

Alpha en 1.34, beta habilitado por defecto en 1.35. Y el detalle que lo hace adoptable de verdad: sigue siendo YAML válido, así que un kubectl viejo lo consume igual. No hace falta reescribir nada a mano — tenés kubectl -o kyaml, el yamlfmt de Kubernetes y el conversor de Google. No es el formato por defecto, así que se adopta de a poco conviviendo con el YAML de siempre.

04 Sep 2026
infoq.com →
☁️

Cloud & DevOps

Patch

Prometheus 3.13.3: un regex case-insensitive te estaba descartando valores en silencio

El bugfix que más importa: los label matchers con regex case-insensitive en PromQL descartaban valores que sí matcheaban. Traducido al turno de guardia: una query que creías que cubría todo tu fleet devolvía menos series de las que hay, y una alerta basada en ese matcher simplemente no dispara. Los peores bugs de observabilidad no son los que rompen el dashboard, son los que te lo dejan verde.

Vienen varios más del mismo calibre: queries out-of-order bloqueando la compaction por horas con la memoria creciendo, series borradas que causaban samples faltantes y errores tras un restart, 100% de CPU en shutdown tanto en el scrape manager como en alerting —lo que demoraba el graceful shutdown hasta gatillar kills por timeout—, y leaks de goroutines y file handles al fallar al abrir un TSDB corrupto (en Windows el handle filtrado además impedía borrar el directorio). Sumale bumps de seguridad de klauspost/compress y golang.org/x/crypto.

07 Sep 2026
github.com/prometheus →
Destacado

Azure APIM: zone redundancy baja a Standard v2, pero no la podés activar en lo que ya tenés

La redundancia de zona deja de ser exclusiva de Premium: Standard v2 la soporta a $700/mes por unidad, contra $2.801 de Premium v2. Diferencia de SLA: 99,95% en Standard v2 contra 99,99% en Premium — aunque ese 99,99% exige desplegar en al menos dos zonas o regiones.

Ahora, la limitación que te arruina el plan: no se puede habilitar sobre una instancia Standard v2 existente. Hay que crear un deployment nuevo, con re-import de APIs, reconfiguración de policies, cambios de DNS y migración de subscription keys. Y si venís bajando de Premium, perdés multi-region, self-hosted gateways, VNET injection y workspaces, más el tope de 50 millones de requests mensuales contra ilimitado.

O sea: el ahorro es real pero el costo de migración también. Hacé la cuenta completa antes de festejar el número de la etiqueta.

07 Sep 2026
infoq.com →
Release

workerd v1.20260908.1: release de build, sin cambios de API

Día tranquilo en el runtime de Workers: tres PRs, todos de build. Optimizaciones internas de JSG —keyear los wrappers de campos por tipo en vez de por campo, y guardarlos en un array plano— más un fix del build interno para forks de miembros de la org. Nada que te toque el código. Lo anotamos para mantener la trazabilidad de la cadencia diaria.

08 Sep 2026
github.com/cloudflare →
🔗

Fullstack

Beta

Vite 8.3.0-beta.1: devtools integrados al dev server

La feature del beta es la integración de devtools con el dev server. El resto son fixes de build que valen la pena si te mordieron: los placeholders de hash ahora se mantienen tal cual en el hook resolveFileUrl, ya no se inlinean los targets de preload link (bug abierto desde el #13355), y findNearestMainPackageData resuelve el root real del package cuando hay package.json anidados — clásico dolor de monorepo.

07 Sep 2026
github.com/vitejs →
Patch

PocketBase 0.40.3: cascade delete de relaciones auto-referenciadas y bugs reportados por bots

Arreglan el cascade delete anidado de records con relaciones auto-referenciadas —si modelás árboles o jerarquías en PocketBase, ese es tuyo—, el validador de índices ahora acepta expresiones con paréntesis en el WHERE opcional, y el validador del campo json se alineó con la semántica de encoding/json/v2 permitiendo keys duplicadas al marshalizar para no romper datos viejos de jsonv1.

Detalle simpático del changelog: el maintainer agrupa varios de estos fixes como reportados por «2 usuarios LLM (¿bots?)». Señal de los tiempos — y recordatorio de que el reporte automatizado encuentra cosas reales, aunque haya que filtrar el ruido.

06 Sep 2026
github.com/pocketbase →
Beta

TanStack Solid Start 2.0.0-rc.7: respeta el server flag en SSR de desarrollo

Un solo fix en el RC: solid-router ahora respeta el flag de server durante el SSR de desarrollo. Sigue avanzando el ciclo de RCs de la serie 2.0 —van siete— así que el stack de TanStack fuera de React se va acercando, de a un fix por vez.

08 Sep 2026
github.com/TanStack →
🔥

Top Stories

Notable

Netflix drops its own autoscaler for Apache Flink's: $1.1M less per year

Netflix's proprietary autoscaler scaled at cluster level: it applied one decision to every operator in a job. That works until you have pipelines with branches, joins, and terabytes of state, where each component has different processing needs. Apache Flink's autoscaler analyzes vertex by vertex, computing each one's required parallelism from throughput and busy time.

The numbers: one team cut 58% of annualized Flink compute spend, roughly $1.1 million a year. Netflix runs over 30,000 streaming jobs across multiple AWS regions.

Two engineering details worth more than the headline. First, they integrated the autoscaler with their control plane using Spring Boot and Temporal workflows to isolate scaling decisions per job, and had to improve JobManager metric collection to support jobs with up to 3,000 subtasks, adding server-side metric filtering. Second, they run a utilization target of 0.45 instead of the community default 0.7, precisely to avoid aggressively rescaling large stateful applications — moving terabytes of state costs more than the capacity you save.

07 Sep 2026
infoq.com →
Notable

Airbnb moves authentication server-side and deletes 60% of the code

The idea is easy to state and hard to accept: instead of the client deciding which authentication method to show, the server makes that decision and the client renders the screen the server returns. That lets them prefer WhatsApp OTP in Brazil or local identity providers in South Korea without shipping new client code.

The 60% code reduction comes from centralization: web, iOS, and Android render standardized server-defined screens instead of embedding the logic, and none of them duplicates challenge selection. On top of that, server-side schemas generate the client type definitions. Result: 100 KB off the web bundle and roughly 11% savings in OTP costs.

The product numbers are strong: experimentation went from weeks to days (20+ experiments in three months), 2.6% improvement in authentication success, and 27% fewer duplicate accounts.

Now, the tradeoffs are real and you should not hide them: more server-side complexity managing policy engines and regional logic, and the client now depends on server availability at every step of login. That last one is not minor — you have become a single point of failure for your product's front door. If you go this way, degradation design is part of the design, not an extra.

04 Sep 2026
infoq.com →
Beta

KYAML: a strict YAML subset so you stop guessing types in Kubernetes

KYAML is a strict subset of YAML: objects use {}, arrays use [], and string values require double quotes. It keeps what is useful about YAML — comments — and strips the ambiguity: indentation-based structure that breaks on its own, and unquoted values reinterpreted as another type. That problem compounds when manifests are generated by a templating system like Helm, which is where it hurts most.

Alpha in 1.34, beta enabled by default in 1.35. And the detail that makes it genuinely adoptable: it is still valid YAML, so an older kubectl consumes it fine. No manual rewriting needed — you have kubectl -o kyaml, Kubernetes' yamlfmt, and Google's converter. It is not the default format, so you adopt it gradually alongside conventional YAML.

04 Sep 2026
infoq.com →
☁️

Cloud & DevOps

Patch

Prometheus 3.13.3: a case-insensitive regex was silently dropping your values

The bugfix that matters most: case-insensitive regex label matchers in PromQL were silently dropping matching values. Translated to your on-call shift: a query you believed covered your whole fleet returned fewer series than exist, and an alert built on that matcher simply does not fire. The worst observability bugs are not the ones that break the dashboard — they are the ones that leave it green.

Several more of the same caliber: out-of-order queries blocking compaction for hours while memory grew, deleted series causing missing samples and errors after restart, 100% CPU on shutdown in both the scrape manager and alerting — delaying graceful shutdown until timeout-based kills triggered — and goroutine and file handle leaks when opening a corrupt TSDB failed (on Windows the leaked handle also prevented directory removal). Plus security bumps for klauspost/compress and golang.org/x/crypto.

07 Sep 2026
github.com/prometheus →
Notable

Azure APIM: zone redundancy comes to Standard v2, but not on what you already run

Zone redundancy stops being Premium-only: Standard v2 supports it at $700/month per unit versus $2,801 for Premium v2. SLA difference: 99.95% on Standard v2 against 99.99% on Premium — though that 99.99% requires deploying across at least two zones or regions.

Now the limitation that wrecks the plan: it cannot be enabled on an existing Standard v2 instance. You must create a new deployment, with API re-imports, policy reconfiguration, DNS changes, and subscription key migration. And if you are stepping down from Premium, you lose multi-region, self-hosted gateways, VNET injection, and workspaces, plus you cap at 50 million monthly requests versus unlimited.

So: the saving is real and so is the migration cost. Run the full sum before celebrating the sticker number.

07 Sep 2026
infoq.com →
Release

workerd v1.20260908.1: a build release, no API changes

A quiet day in the Workers runtime: three PRs, all build-related. Internal JSG optimizations — keying struct field wrappers on the field type rather than the field, and storing them in a flat array — plus an internal build fix for forks of org members. Nothing that touches your code. Noted to keep the daily cadence traceable.

08 Sep 2026
github.com/cloudflare →
🔗

Fullstack

Beta

Vite 8.3.0-beta.1: devtools integrated into the dev server

The beta's feature is wiring devtools into the dev server. The rest are build fixes worth knowing if they bit you: hash placeholders now stay as-is in the resolveFileUrl hook, preload link targets are no longer inlined (an issue open since #13355), and findNearestMainPackageData resolves the actual package root with nested package.json files — a classic monorepo pain.

07 Sep 2026
github.com/vitejs →
Patch

PocketBase 0.40.3: self-referenced cascade deletes, and bugs reported by bots

They fix nested cascade delete of self-referenced relation records — if you model trees or hierarchies in PocketBase, that one is yours — the index validator now accepts expressions with parentheses in the optional WHERE clause, and the json field validator was aligned with encoding/json/v2 semantics, allowing duplicate keys on record marshaling so old jsonv1 data keeps working.

A nice changelog detail: the maintainer groups several of these fixes as reported by «2 LLM users (bots?)». Sign of the times — and a reminder that automated reporting does find real things, even if you have to filter the noise.

06 Sep 2026
github.com/pocketbase →
Beta

TanStack Solid Start 2.0.0-rc.7: respects the server flag in development SSR

A single fix in this RC: solid-router now respects the server flag during development SSR. The 2.0 RC cycle keeps moving — seven and counting — so TanStack's non-React stack inches closer, one fix at a time.

08 Sep 2026
github.com/TanStack →
🔥

Top Stories

Notable

Netflix abandonne son autoscaler pour celui d'Apache Flink : 1,1 M$ de moins par an

Netflix's proprietary autoscaler scaled at cluster level: it applied one decision to every operator in a job. That works until you have pipelines with branches, joins, and terabytes of state, where each component has different processing needs. Apache Flink's autoscaler analyzes vertex by vertex, computing each one's required parallelism from throughput and busy time.

The numbers: one team cut 58% of annualized Flink compute spend, roughly $1.1 million a year. Netflix runs over 30,000 streaming jobs across multiple AWS regions.

Two engineering details worth more than the headline. First, they integrated the autoscaler with their control plane using Spring Boot and Temporal workflows to isolate scaling decisions per job, and had to improve JobManager metric collection to support jobs with up to 3,000 subtasks, adding server-side metric filtering. Second, they run a utilization target of 0.45 instead of the community default 0.7, precisely to avoid aggressively rescaling large stateful applications — moving terabytes of state costs more than the capacity you save.

07 Sep 2026
infoq.com →
Notable

Airbnb déplace l'authentification côté serveur et supprime 60% du code

The idea is easy to state and hard to accept: instead of the client deciding which authentication method to show, the server makes that decision and the client renders the screen the server returns. That lets them prefer WhatsApp OTP in Brazil or local identity providers in South Korea without shipping new client code.

The 60% code reduction comes from centralization: web, iOS, and Android render standardized server-defined screens instead of embedding the logic, and none of them duplicates challenge selection. On top of that, server-side schemas generate the client type definitions. Result: 100 KB off the web bundle and roughly 11% savings in OTP costs.

The product numbers are strong: experimentation went from weeks to days (20+ experiments in three months), 2.6% improvement in authentication success, and 27% fewer duplicate accounts.

Now, the tradeoffs are real and you should not hide them: more server-side complexity managing policy engines and regional logic, and the client now depends on server availability at every step of login. That last one is not minor — you have become a single point of failure for your product's front door. If you go this way, degradation design is part of the design, not an extra.

04 Sep 2026
infoq.com →
Bêta

KYAML : un sous-ensemble strict de YAML pour cesser de deviner les types

KYAML is a strict subset of YAML: objects use {}, arrays use [], and string values require double quotes. It keeps what is useful about YAML — comments — and strips the ambiguity: indentation-based structure that breaks on its own, and unquoted values reinterpreted as another type. That problem compounds when manifests are generated by a templating system like Helm, which is where it hurts most.

Alpha in 1.34, beta enabled by default in 1.35. And the detail that makes it genuinely adoptable: it is still valid YAML, so an older kubectl consumes it fine. No manual rewriting needed — you have kubectl -o kyaml, Kubernetes' yamlfmt, and Google's converter. It is not the default format, so you adopt it gradually alongside conventional YAML.

04 Sep 2026
infoq.com →
☁️

Cloud & DevOps

Patch

Prometheus 3.13.3 : un regex insensible à la casse ignorait vos valeurs en silence

The bugfix that matters most: case-insensitive regex label matchers in PromQL were silently dropping matching values. Translated to your on-call shift: a query you believed covered your whole fleet returned fewer series than exist, and an alert built on that matcher simply does not fire. The worst observability bugs are not the ones that break the dashboard — they are the ones that leave it green.

Several more of the same caliber: out-of-order queries blocking compaction for hours while memory grew, deleted series causing missing samples and errors after restart, 100% CPU on shutdown in both the scrape manager and alerting — delaying graceful shutdown until timeout-based kills triggered — and goroutine and file handle leaks when opening a corrupt TSDB failed (on Windows the leaked handle also prevented directory removal). Plus security bumps for klauspost/compress and golang.org/x/crypto.

07 Sep 2026
github.com/prometheus →
Notable

Azure APIM : la zone redundancy arrive en Standard v2, mais pas sur l'existant

Zone redundancy stops being Premium-only: Standard v2 supports it at $700/month per unit versus $2,801 for Premium v2. SLA difference: 99.95% on Standard v2 against 99.99% on Premium — though that 99.99% requires deploying across at least two zones or regions.

Now the limitation that wrecks the plan: it cannot be enabled on an existing Standard v2 instance. You must create a new deployment, with API re-imports, policy reconfiguration, DNS changes, and subscription key migration. And if you are stepping down from Premium, you lose multi-region, self-hosted gateways, VNET injection, and workspaces, plus you cap at 50 million monthly requests versus unlimited.

So: the saving is real and so is the migration cost. Run the full sum before celebrating the sticker number.

07 Sep 2026
infoq.com →
Release

workerd v1.20260908.1 : release de build, sans changement d'API

A quiet day in the Workers runtime: three PRs, all build-related. Internal JSG optimizations — keying struct field wrappers on the field type rather than the field, and storing them in a flat array — plus an internal build fix for forks of org members. Nothing that touches your code. Noted to keep the daily cadence traceable.

08 Sep 2026
github.com/cloudflare →
🔗

Fullstack

Bêta

Vite 8.3.0-beta.1 : devtools intégrés au dev server

The beta's feature is wiring devtools into the dev server. The rest are build fixes worth knowing if they bit you: hash placeholders now stay as-is in the resolveFileUrl hook, preload link targets are no longer inlined (an issue open since #13355), and findNearestMainPackageData resolves the actual package root with nested package.json files — a classic monorepo pain.

07 Sep 2026
github.com/vitejs →
Patch

PocketBase 0.40.3 : cascade delete auto-référencé et bugs signalés par des bots

They fix nested cascade delete of self-referenced relation records — if you model trees or hierarchies in PocketBase, that one is yours — the index validator now accepts expressions with parentheses in the optional WHERE clause, and the json field validator was aligned with encoding/json/v2 semantics, allowing duplicate keys on record marshaling so old jsonv1 data keeps working.

A nice changelog detail: the maintainer groups several of these fixes as reported by «2 LLM users (bots?)». Sign of the times — and a reminder that automated reporting does find real things, even if you have to filter the noise.

06 Sep 2026
github.com/pocketbase →
Bêta

TanStack Solid Start 2.0.0-rc.7 : respecte le server flag en SSR de développement

A single fix in this RC: solid-router now respects the server flag during development SSR. The 2.0 RC cycle keeps moving — seven and counting — so TanStack's non-React stack inches closer, one fix at a time.

08 Sep 2026
github.com/TanStack →