Page 16 - Mironsoft Blog

In many organically grown Magento databases, the entire application, every cron job, and every reporting tool connects through the same, often years-old database account with full privileges. This article shows how to enforce defensible password policies with the validate_password component, configure password expiration and history cleanly, and apply the principle of least privilege with clearly separated accounts for read-only reporting, migrations, and actual application runtime.

When PHP-FPM or Nginx crashes under load with Too many open files, the cause is almost always a ulimit value set too low, not a genuine resource shortage. This article explains the difference between soft and hard limits, shows how to configure file descriptor and process boundaries persistently and correctly through limits.conf and systemd overrides, and how to reliably diagnose the error.

Since TypeScript 5.0, type parameters can carry the const modifier. This makes the compiler automatically infer the narrowest possible literal types on every call, without the caller having to write as const.

Bash can open TCP sockets, build HTTP requests by hand, and test ports without curl, wget, or netcat, through the built-in /dev/tcp pseudo-device. Understanding the mechanics helps you secure minimal containers and restrictive environments, but you also need to know the clear limits: no TLS, no UDP with real feedback, and no header control like a real HTTP client offers.

When a single Redis instance in Magento reaches its limits, two paths open up: a real Redis Cluster with sharding, or a clean separation of cache, session and full page cache onto dedicated Redis instances. This article shows when each path makes sense and how the transition actually works.

Browsers constantly make decisions about which resource to request next while loading a page. The fetchpriority attribute lets you override those heuristics on purpose and tell the browser exactly which images, scripts, or links are actually critical for the first paint.

Every Docker image starts with a base image that the whole team trusts blindly, usually referenced by a tag like latest or a version number without further checks. Supply chain security replaces that blind trust with pinned digests, cryptographic signatures, and traceable provenance information, so a compromised base image does not silently reach production deployments.

Anyone who only uses git log with its default output is giving up one of the most powerful diagnostic tools in daily development work. With targeted filters by author, date and content, with custom pretty formats for scripts, and with pickaxe search, any line of code can be traced back to its origin without manually scanning the history.

Every full text search in Elasticsearch stands or falls with the analyzer that breaks text into comparable tokens at index time and at search time. Understanding character filters, tokenizers and token filters as a pipeline lets you build your own analyzers and understand why a search returns hits, or does not.

Not every component needs Server-Sent Events or WebSocket, a simple but cleanly implemented polling loop is often entirely sufficient. The problem is rarely the basic idea behind setInterval, it is the details around it: an interval that keeps running after the component gets removed, a server that keeps getting hit every second while it is already down, or a background tab that happily keeps polling even though nobody is looking. This article shows how to tie a polling interval cleanly to Alpine's init() and destroy() lifecycle, how exponential backoff kicks in on repeated failures, and how polling can be paused through the Page Visibility API as soon as the tab is no longer visible.