← All stories
● Covered by 1 source · 1 reportMedium impact1 negative

AWS Lambda Web Functions Bill Memory for 5-7 Minutes After Last Request

🔄 Updated 9d ago
New to BrevFeed? We gather this story from every outlet covering it into one summary — ranked by real-world impact, not just the latest headline — so you never miss what matters. What is BrevFeed? →

Key points

  • Lambda Web Functions bill memory for 5-7 minutes after the last request.
  • AWS documentation does not specify idle environment termination duration.
  • Idle memory charges accounted for 88% of compute costs in tests.
  • Custom compute infrastructure was faster and more cost-effective.

Unexpected Idle Billing Behavior

AWS Lambda Web Functions were observed to bill for memory for 5-7 minutes after an environment's last request, even during periods of inactivity. This behavior was discovered during testing by Easel, a company that initially considered using Web Functions to replace custom infrastructure for running web applications.

While AWS documentation states that memory is billed throughout an execution environment's lifetime, it does not specify the duration an idle environment remains active before termination. Easel's tests measured idle periods ranging from 313 to 411 seconds across representative environments, with no configurable idle timeout option available.

Cost Implications for Infrequent Workloads

For applications with infrequent or unpredictable traffic, this extended idle billing period significantly impacts costs. In Easel's experiments, a 200ms request followed by 345 seconds of inactivity resulted in memory being metered for approximately 1,700 times the request duration.

Analysis using CloudWatch and AWS Cost Explorer showed that memory charges constituted about 88% of the total compute and request costs, even though the application only used 160-210 MB of memory. While the total cost for the experiment was small, this billing model becomes problematic for thousands of applications with sporadic usage.

Performance and Cost Comparison

Due to the billing issue, Easel opted to build its own compute infrastructure. This custom solution not only addressed the cost concerns but also delivered better performance. The custom compute answered warm requests 1.5-3 times faster and cold-started approximately 3.5 times faster than AWS Lambda Web Functions for the same Next.js application.

Impact on Application Architectures

This finding suggests that while AWS Lambda Web Functions may be suitable for applications with steady traffic where idle costs are amortized across many requests, they present a significant cost inefficiency for workloads characterized by many idle periods between requests. This could influence architectural decisions for companies managing numerous applications with varying traffic patterns.

✨ This summary was generated by AI from the outlets' reporting listed below. It is not independently verified and may contain errors — check the original sources. How BrevFeed works →

The daily brief

One email each morning: the day's tech stories, clustered across outlets and summarized. No account needed.

One email a day. Unsubscribe in one click, any time.

Today's brief

Spend a few minutes, get the whole day. Every topic's top stories in one hands-free rundown — listen, watch, or read the transcript.

~4 min · 3 stories · Nov 21

▶ Play today's brief Listen on Spotify

New every morning, and the back catalogue is archived by date.

Reporting from

AWS Lambda Web Functions continue to bill for memory for 5-7 minutes after the last request, even when idle, a behavior not specified in documentation regarding idle environment termination time. This extended billing period significantly increases costs for applications with infrequent or unpredictable traffic, making custom compute infrastructure more cost-effective and performant for certain workloads.