This week in caching: hardening a Redis cache box before prod

A pre-flight checklist so your cache layer doesn't become an incident. Skip if you're on a managed service.

— Pick the right eviction policy — a pure cache wants allkeys-lru or allkeys-lfu; the default noeviction turns a full cache into write errors across your app.
— Set maxmemory explicitly — unset, Redis eats RAM until the OOM killer takes it down at peak.
— Disable dangerous commands — rename or block FLUSHALL, KEYS, and CONFIG in production; one stray KEYS * blocks the whole server.
— Bind and auth — never expose 6379 to the internet; set requirepass and bind to private interfaces. Open Redis boxes get cryptomined within hours.
— Watch evicted_keys and used_memory — rising evictions mean your working set outgrew memory; scale before the hit ratio collapses.

Credit to the Redis admin and security docs for the rename-command hardening.

Bookmark: the Redis 'Securing Redis' page — the eviction + auth defaults everyone should change.

This week in caching: Redis round-trip reduction

The cache was fast; the network wasn't. How teams fixed the gap:

— Feed service — replacing 300 sequential GETs with one MGET / pipeline dropped page render from 240ms to 18ms; same Redis, the win was killing 300 network round-trips.
— Laravel app — moving Redis off a remote managed host onto a local unix socket cut per-call latency from 0.8ms to 0.05ms; on a page doing 200 cache calls that's a real 150ms.
— API — Lua scripting a read-modify-write into one atomic call removed a race and a round-trip at once.

The lesson: cache hit latency is dominated by network round-trips, not Redis itself. Batch and co-locate.

Bookmark: redis-cli --latency and --intrinsic-latency — separate network cost from server cost before you tune.

Microsoft планирует вставлять рекламу в игры

➡️ Читайте на сайте: https://aff.top/blog/microsoft-planiruet-vstavliat-reklamu-v-igry

🧠 Ещё больше инсайтов → в канале AFF.top

Myth: more cache layers always means faster
This week in caching, the layering fallacy:
— The claim: stack page cache + object cache + CDN + browser cache and speed compounds.
— The correction: each layer adds an invalidation surface; a stale object cache can poison your fresh page cache, so a miss now costs a multi-layer rebuild instead of one.
— Worth your time: audit which layer actually owns the TTL for a given URL before adding another. Two coherent layers beat four that fight over the same key.
Bookmark: map your stack as a dependency tree, not a pile. The bottom layer's staleness window caps everything above it.

This week in caching: private browser cache vs shared CDN cache

The header nuance that causes the most cache bugs:

— Cache-Control: private means only the user's browser may store it — correct for logged-in dashboards. A CDN must skip it.
— public opens it to shared caches (CDN, proxies) — correct for anonymous pages.
— The killer: s-maxage overrides max-age for shared caches only. So you can tell the CDN 'cache 1 hour' while telling browsers 'cache 1 minute' — long edge life, fresh-ish clients, no header conflict.

Classic outage: serving a logged-in page as public and leaking one user's cart to everyone behind the CDN. Always pair user-specific responses with private AND a Vary on the auth cookie.

Bookmark: MDN's Cache-Control reference — keep the private/public/s-maxage section open while you debug.

Myth: set opcache.revalidate_freq=0 for safety
This week in caching, the OPcache stat trap:
— The advice you've seen: keep revalidate_freq at 0 so PHP always picks up code changes.
— Why it's wrong: 0 means a filesystem stat() on every include on every request, which is the exact syscall overhead OPcache exists to kill.
— The real move: on production set opcache.validate_timestamps=0 and reload PHP-FPM on deploy. That's the configuration that actually buys you the cached-bytecode win.
Bookmark: revalidate_freq only matters when validate_timestamps is on; turning the latter off makes the former irrelevant. Tie cache busting to your deploy, not to the clock.

This week in caching: Vary header vs separate URLs for variants

Serving mobile/desktop or multi-language from one cache — two roads:

— Vary header (e.g. Vary: Accept-Encoding) keeps one URL and lets the cache store per-variant. Clean for encoding/compression. But Vary: User-Agent is a hit-rate killer — thousands of UA strings = thousands of cache entries for one page.
— Separate URLs (/en/, /de/, m.site.com): each variant is its own cacheable object with high hit rate, and it's SEO-friendly. Cost: routing/redirect logic and canonical tags.
— The middle path: normalize the varying input to a small set (UA → 'mobile'/'desktop' bucket) before it hits the cache key, so you get Vary's single-URL simplicity without the explosion.

Rule: Vary only on low-cardinality headers; for anything high-cardinality, bucket it or split the URL.

Bookmark: Smashing Magazine's 'Vary header' deep-dive on cache-key cardinality.

На Anthropic, OpenAI, Google и SpaceXAI подали в суд из-за ИИ

В Калифорнии против ИИ-компаний подали антимонопольный иск: регулятору показалось подозрительным, что игроки синхронно призывают ограничить развитие нейросетей ради безопасности. Смысл спора в том, что инвестиции в ИИ уже обгоняют реальный прогресс, а бизнесу выгодны правила, которые защитят капитал. Вывод: быстрых прорывов ждать не стоит, лучше выжимать максимум из текущих инструментов.

➡️ Читайте на сайте: https://aff.top/blog/na-anthropic-openai-google-i-spacexai-podali-v-sud-iz-za-ii

🧠 Ещё больше инсайтов → в канале AFF.top

Google ads начал показывать расходы конкурентов

Google Ads запустил Peer Spending — инструмент, который сравнивает расходы аккаунта с рекламодателями из той же ниши без раскрытия чужих данных. Он показывает, тратите вы больше, меньше или примерно на уровне конкурентов на уровне кампаний и групп объявлений. Для арбитража это скорее ориентир по бенчмаркам, чем инструмент прямого усиления залива.

➡️ Читайте на сайте: https://aff.top/blog/google-ads-nachal-pokazyvat-raskhody-konkurentov

🧠 Ещё больше инсайтов → в канале AFF.top

This week in caching: moving PHP sessions to Redis the safe way

A migration checklist that avoids logging everyone out. Skip if you're stateless via JWT.

— Separate session DB from cache DB — use a different Redis numbered DB (or instance) so a FLUSHDB of your cache never dumps live sessions.
— Set persistence honestly — sessions need AOF or RDB; a pure-cache Redis with no persistence loses every login on restart.
— Match TTL to session.gc_maxlifetime — let Redis expire keys instead of PHP's GC probability lottery.
— Lock carefully — the Redis session handler locks per-session; long-running AJAX can serialize requests and feel like a hang. Set session.lock_expire.
— Drain, don't cut over — run both handlers during deploy or accept a mass logout.

Credit to Colin Mollenhour's php-redis-session-abstract docs for the locking gotchas.

Bookmark: the PHP session.save_handler=redis ini reference — the exact directives.

Myth: a CDN means you don't need page caching
This week in caching, the edge overconfidence file:
— The pitch: put Cloudflare in front and origin caching becomes redundant.
— The catch: CDNs cache by default only for static extensions; your HTML usually passes through unless you explicitly set Cache-Control and a page rule. A logged-in cookie or a query string can silently make everything a MISS to origin.
— Worth your time: check your edge cache HIT ratio on HTML specifically, not on the aggregate that fonts and images inflate.
Bookmark: the edge caches what you tell it to cache. An uncached origin behind a CDN is still an uncached origin for every cold edge node.

Минфин РФ планирует выпустить собственный стейблкоин

Власти РФ обсуждают запуск рублёвого стейблкоина: сейчас решают, как его обеспечить, какие операции разрешить и будет ли на него спрос. Основной кейс — международные переводы, а не использование физлицами. Если проект доведут до запуска, он может стать частью новой криптоинфраструктуры и альтернативой токенам, привязанным к дружественным валютам.

➡️ Читайте на сайте: https://aff.top/blog/minfin-rf-planiruet-vypustit-sobstvennyi-steiblkoin

🧠 Ещё больше инсайтов → в канале AFF.top

В публичный доступ вышел Grok 4.7

➡️ Читайте на сайте: https://aff.top/blog/v-publichnyi-dostup-vyshel-grok-4-7

🧠 Ещё больше инсайтов → в канале AFF.top

Myth: Redis object cache is always faster than disk
This week in caching, the Redis-as-default reflex:
— The assumption: in-memory beats file-based transients every time.
— The nuance: a local Redis over a unix socket, yes; a managed Redis one network hop away can lose to a warm OS page cache on local SSD once round-trip latency stacks across dozens of gets per page.
— Worth your time: count your object-cache calls per request first. 400 gets at 0.4ms network each is 160ms you didn't have on disk.
Bookmark: 'in-memory' only wins when the memory is actually near. Measure get-count x round-trip before you assume Redis is the upgrade. Credit to anyone who's ever profiled a remote Redis under load.

X теперь показывает причину теневого бана

➡️ Читайте на сайте: https://aff.top/blog/x-teper-pokazyvaet-prichinu-tenevogo-bana

🧠 Ещё больше инсайтов → в канале AFF.top

Россиян заставят отчитываться перед ФНС за крипту

Крипта в РФ считается имуществом, поэтому доход от продажи или обмена облагается НДФЛ: 13% до 2,4 млн рублей и 15% сверх этого. Декларацию 3-НДФЛ нужно подать самостоятельно до 30 апреля. Отчитываться по кошелькам обязаны только при обороте свыше 600 000 рублей в год, а за нарушение грозят штрафы и пени.

➡️ Читайте на сайте: https://aff.top/blog/rossiian-zastaviat-otchityvatsia-pered-fns-za-kriptu

🧠 Ещё больше инсайтов → в канале AFF.top

В публичный доступ вышла ChatGPT-6

OpenAI выпустила линейку GPT-6: модель Sol для кодинга и сложных задач и Luna для рутинных операций. Ключевое преимущество релиза — двукратное снижение цен на API ($2 и $0.10 за 1 млн входных токенов соответственно) и сокращение числа ошибок в два раза. Это позволяет масштабировать автоматизацию и разработку с вдвое меньшими затратами на инфраструктуру.

➡️ Читайте на сайте: https://aff.top/blog/v-publichnyi-dostup-vyshla-chatgpt-6

🧠 Ещё больше инсайтов → в канале AFF.top

This week in caching: a cache-warming runbook for launch day

Don't let your first real visitors pay for every cold miss. Run this before you announce.

— Crawl your own sitemap — a simple wget --quiet -i sitemap-urls.txt or a headless crawler populates the page cache ahead of traffic.
— Warm in priority order — homepage, top landing pages, then long tail; if the warm-up gets cut short, the important pages are already hot.
— Pre-render representative variants — if you cache per device or per format, warm each variant, not just desktop.
— Warm object/DB cache too — hit the endpoints that fill Redis, not just the HTML edge; a warm page cache over a cold object cache still stalls on first dynamic call.
— Check Age headers afterward — a populated cache shows a nonzero Age; zero everywhere means your warm-up didn't stick.

Credit to the cache-preload features in WP Rocket and Cloudflare's prefetch guidance.

Bookmark: a tiny sitemap-to-curl warming script — the WP and static crowds both reuse it.