--- title: "RPO (Recovery Point Objective)" type: concept tags: [灾难恢复, 数据保护, 指标] sources: ["https://launchdarkly.com/blog/rto-vs-rpo/"] last_updated: 2025-07-26 --- ## Definition RPO(Recovery Point Objective,恢复点目标)是指可接受的最大数据丢失量,以时间度量。从最后一次有效备份到故障发生所经历的时间。 ## Key Characteristics - 衡量数据保护程度 - 从故障时刻向后计算(倒计时) - 与备份频率直接相关 - 需要与 RTO 共同规划,不能只优化其中一个 ## Tiered RPO Targets (from this source) | Tier | Examples | RPO Target | |------|----------|------------| | Critical | Payment processing, user auth | < 1 minute | | Important | Admin dashboards, reporting | < 15 minutes | | Nice-to-have | Internal tools, dev environments | < 1 hour | ## Example 如果数据库在下午3点崩溃,而最后一次备份是下午2点,则 RPO 为1小时。2点到3点之间的所有数据丢失。 ## Connections - [[RTO (Recovery Time Objective)]] ← 配对指标 → [[RPO (Recovery Point Objective)]] - [[灾难恢复]] ← 应用领域 → [[RPO (Recovery Point Objective)]] - [[持续交付]] ← 现代上下文 → [[RPO (Recovery Point Objective)]] - [[Feature Flag]] ← 保护工具 → [[RPO (Recovery Point Objective)]]