Medições — os relatórios completos por trás dos números

Toda afirmação de desempenho e segurança deste site vem de uma rodada que foi executada, registrada e documentada. Estes são esses relatórios na íntegra: o hardware, o conjunto de dados, as condições, os números brutos e as afirmações que não resistiram à atribuição e foram retiradas. Várias rodadas concluíram que uma mudança que esperávamos ser importante não fez diferença mensurável. Elas também estão aqui.

The aaPanel benchmark, in fullSixPanel 1.4 — the same test, run again on 28 September 2026 · Two bugs in our own benchmark, found before the numbers were used · Single runs lie on machines this size
MariaDB versions, measured — four pairs, no reportable differenceThe short answer · The first database comparison was invalid, and the floor is what gave it away · The version pairs, measured properly
The OS and stack round — six axes, one reportable resultThe short answer · The six axes, and what each returned · The noise floor, which limits every other number on this page
PHP versions, measured — four pairings, four washesThe short answer · The four pairings · The three-way, in full
Processor architecture, measured — arm64 and x86_64, and no winnerThe short answer · The boxes · Processor — seconds for a fixed job, median of 5
Security and isolation, measuredWhat was actually attempted, not asserted · The leak side and the attack side are different questions · What the kernel namespace closed, and what it cost
How the 6amMart optimisation was measuredThree kinds of number, and why they stay apart · The admin sweep — all 297 pages · The acceptance rule
What we withdrewThe measurement that forced the rest · "+32 % from the distribution's own packages" · "MariaDB 11.x is 2.7× slower"
How we measureThe rules, in short · The noise floor, and why it dictates everything else · Control rows are the resolution floor
Websites and apps, measured — six kinds on one serverWhat was deployed · Time to answer, at the server · What was checked beyond the timing