发表机构
DeepSeek（深度求索）
机构由 AI 辅助整理，请以论文原文为准。AI 中文总结
DSec是一个生产级弹性沙箱平台，通过统一SDK提供多种后端，支持大规模智能体训练，实现高密度执行与状态保留，每天服务约300万沙箱。
AI 中文摘要
大规模语言模型（LLMs）的智能体训练与评估依赖于隔离的、有状态的执行环境，模型在其中检查代码仓库、调用工具、执行命令，并与特定任务的服务进行交互。这些工作负载会以大规模突发方式创建沙箱，涵盖异构功能与隔离需求，在长时间交互中保持状态，并从复用率较低的大型镜像库中获取资源。因此，支持这些工作负载需要一个弹性执行平台，而非单一的沙箱运行时。本报告介绍了DeepSeek弹性计算（DSec），一个生产级沙箱平台，通过统一的SDK提供FnCall、容器、微虚拟机（microVM）和全虚拟机（full-VM）四种沙箱后端。DSec在集群范围内协调沙箱的放置与生命周期管理，从独立版本化的层中组合环境，结合内存共享、回收与CPU调度以实现高密度执行，并按需从Fire-Flyer文件系统（3FS）——一个集群范围的分布式文件系统——加载镜像数据。DSec与强化学习（RL）框架协同设计，将有状态的推演（rollout）执行与可抢占的GPU训练解耦，将沙箱生命周期与训练协调，以在回收空闲资源的同时保留推演状态，并缓解智能体的不当行为（如奖励黑客攻击）。DSec的一个生产级单元涵盖约160个节点，每天服务约300万个沙箱；在生产环境中，它支持超过38万个并发沙箱，并持续保持每秒超过5000个沙箱的创建速率。我们的评估与部署经验表明，这些机制降低了环境搭建和镜像分发开销，提高了内存效率，并在高密度超卖下保持了延迟敏感性能。
英文摘要
Large-scale agentic training and evaluation with large language models (LLMs) rely on isolated, stateful execution environments in which models inspect repositories, invoke tools, execute commands, and interact with task-specific services. These workloads create sandboxes in large bursts, span heterogeneous functionality and isolation requirements, retain state across long interactions, and draw from large image corpora with limited reuse. Supporting them therefore requires an elastic execution platform rather than a single sandbox runtime. This report presents DeepSeek Elastic Compute (DSec), a production sandbox platform that exposes FnCall, container, microVM, and full-VM sandbox backends through a unified SDK. DSec coordinates placement and lifecycle management across the cluster, composes environments from independently versioned layers, combines memory sharing, reclamation, and CPU scheduling for high-density execution, and loads image data on demand from Fire-Flyer File System (3FS), a cluster-wide distributed filesystem. DSec is co-designed with the reinforcement learning (RL) framework, decouples stateful rollout execution from preemptible GPU training, coordinates sandbox lifecycle with training to preserve rollout state while reclaiming idle resources, and mitigates agent misbehavior such as reward hacking. A single production-scale unit of DSec spans around 160 nodes, serving about 3 million sandboxes per day; in production, it supports over 380,000 concurrent sandboxes and sustains over 5,000 sandbox creations per second. Our evaluation and deployment experience show that these mechanisms reduce environment setup and image-distribution overhead, improve memory efficiency, and preserve latency-sensitive performance under high-density overcommit.
Comments31 pages, 13 figures. This version has been substantially expanded from an earlier version, whose two-page extended abstract underwent first-round review for the Operational Systems Track of ACM SIGOPS ATC 2026
