Content Deep Dive

Sandbox resourcing: burst, don't reserve

Blog post from Modal

Post Details
Company
Date Published
Author
-
Word Count
5,568
Company Posts That Month
6
Language
English
Hacker News Points
-
Post removed?
No
Summary

Modal introduces active-resource pricing and automatic bursting for Sandboxes, arguing that coding agents, RL rollouts, evaluations, and app-building workloads are typically idle for much of their runtime but briefly require substantial CPU or memory. Instead of reserving and paying continuously for peak capacity, Sandboxes begin with a default guaranteed request of 0.125 CPU cores and 128 MiB of memory, then can temporarily use additional shared machine capacity and are billed each second for the higher of their request or actual use. An example coding-agent session spent most of its nearly three-minute duration waiting for model responses or timers while using little CPU and about 47 MiB of memory, despite short testing spikes. Modal estimates that a representative 20-minute coding task could cost 58% less under usage-based bursting than under a fixed reservation sized for peak demand. Users can set higher requests for guaranteed resources or limits to ensure consistent benchmark environments and cap costs, though memory limits can terminate workloads that exceed them. If shared capacity becomes constrained, bursting CPU workloads may slow down, while Sandboxes exceeding their memory request may rarely be evicted; Modal states an OOM-rate objective for bursting Sandboxes below 8 GiB and provides snapshots, volumes, and termination-reason tools to support recovery.

Trends Found in this Post

No tracked trend matches for this post yet.

Use This Data

Use this post, company, and trend context to find content marketing opportunities, perform competitive analysis, or address product feature gaps via the Plushcap MCP server or the Plushcap API.