Your loop yields every 500 items. It still freezes. Here's why (and a fix)

说明与用途

每 500 项循环一次的做法在出现一批重项时仍会导致事件循环冻结。time‑sliced 按时间切片而非计数进行 yield，保持事件循环响应性并保持总运行时间相近。

问题

Long loops that yield every N items can still freeze the event loop when a cluster of heavy items appears, causing timers to fire late and requests to time out.

结果

time‑sliced achieves similar total runtime to a plain loop while keeping the longest event‑loop freeze down to 11 ms (vs 254 ms with every‑500 strategy) and remaining responsive during large imports.

工作流

• Process rows one by one
• Measure elapsed time since last yield
• If elapsed time exceeds budget, await yieldToEventLoop() (setImmediate or setTimeout(0) depending on runtime)
• Reset budget timer and continue processing

证据

When items are cheap, you yield rarely, so overhead stays low. When items are slow, you yield often, so nothing freezes.

来源

time‑sliced finishes in about the same time as the plain loop, and the event loop never freezes much longer than the budget.

来源

A 1 ms heartbeat timer measures the longest event-loop freeze. Median of 3 runs, Node.js 24.

来源