SLA in the evolution from ITSM to SRE: metric transformation and management implementation — evidence from frontline video-networked IoT operations
SLA在ITSM到SRE的指标演化与管理落地中的作用：来自视频联网物联运维的一线证据

摘要

数字化运维从传统的IT服务管理向站点可靠性工程的转型，反映了从流程合规到工程驱动可靠性的范式转变。尽管服务等级协议长期以来一直是服务保障的基石，但新兴的分布式架构和智能联网系统暴露了其局限性。本文研究了可靠性指标从SLA到服务等级目标与错误预算的演化，重点关注其在大规模视频联网和物联网环境中的落地实践。通过整合理论分析与运维洞察，本研究探讨了指标演化如何增强服务韧性、自动化水平和用户体验。结果表明，与传统SLA模型相比，面向SRE的指标系统将故障响应效率提高了35%，并将停机时间方差降低了42%。研究得出结论，在现代IT生态系统中，仅靠SLA已不足够；相反，需要一种整合了量化可靠性工程与自适应反馈机制的混合治理框架，以实现可持续的数字化运营。

Abstract

The transition of digital operations from traditional IT Service Management (ITSM) to Site Reliability Engineering (SRE) reflects a paradigm shift from process compliance to engineering-driven reliability. Although Service Level Agreements (SLAs) have long served as the cornerstone of service assurance, emerging distributed architectures and intelligent networked systems have exposed their limitations. This study examines the evolution of reliability metrics from SLAs to Service Level Objectives (SLOs) and error budgets, with a particular focus on their practical implementation in large-scale video networking and Internet of Things (IoT) environments. By integrating theoretical analysis with operational insights, the research explores how metric evolution enhances service resilience, automation capability, and user experience.
Results indicate that compared with traditional SLA models, SRE-oriented metric systems improve incident response efficiency by 35% and reduce downtime variance by 42%. The study concludes that SLAs alone are no longer sufficient in modern IT ecosystems; instead, a hybrid governance framework integrating quantitative reliability engineering and adaptive feedback mechanisms is required to achieve sustainable digital operations.