for calculating various collection stats by batches of ids, Mongo outcompetes Postgres on larger documents but loses on smaller ones
https://feedx.net，这一点在WhatsApp Web 網頁版登入中也有详细论述
const chunk = new Uint8Array(chunkSize);。谷歌对此有专业解读
seems it's solved! Looking at profilers, we see millions fewer allocations in