Skip to content
NewA calmer, faster ZeroHour
Otherinfo

Lambda Web Functions bill idle memory for about 5–7 minutes

Original title: AWS Lambda Web Functions keep billing memory ~7 minutes after the last request

Hacker News · AIby seanw265 ()5 min read1
AI summary · x-ai/grok-4.7Summaries by Vani AI

Easel measured 313–411 seconds of idle billing and reported faster warm and cold starts on its own hosts.

Easel tested AWS Lambda Web Functions and found execution environments continued billing memory for 5–7 minutes after the last request, with no configurable idle timeout. Across the experiment, memory was about 88% of calculated compute and request cost, while the whole test cost about two cents at list prices. Easel moved application execution to servers it operates and reported faster median warm requests and cold starts than Web Functions on a small Next.js 16 comparison.

Unconfirmed: Easel says the benchmarks are early, from one region, with only two cold-start samples and no sustained-throughput test.

Key facts
  • Six us-east-1 environments idled 313–411 seconds on October 26.
  • A 200ms request plus 345 seconds idle was about 1,700 times the request duration.
  • Exiting after 10 seconds cut metered time about 20× but was logged as a runtime failure.
  • Easel's first self-operated US East host went live on November 2.
  • After 12 minutes idle, two trials showed 216–250 ms versus 810–854 ms.
Explain this AI news with Vani AIWhat was announced, what is new and why it matters, in a new chat

From the original article

5 min read at easel.sh

When AWS announced Lambda Web Functions in October, I was excited. At Easel , we'd spent months building infrastructure to run web applications efficiently on traditional Lambda. Native HTTP servers, concurrent requests, and active CPU billing sounded like exactly what we'd been looking for.

Hacker News · AI

By seanw265 · easel.sh

Read the full article at Hacker News · AI

A short excerpt; the article and its images belong to the publisher. Summary, key facts and assessment above are AI-written from the article and can be wrong.