运维响应标准升级：30 分钟内响应成为默认承诺Upgraded Support Standards: 30-Minute Response Now the Default Commitment

从 2026 年 5 月 1 日起，我们对所有在保项目统一升级运维响应标准：严重故障 30 分钟内响应、4 小时内给出处理方案与预估恢复时间。Effective May 1, 2026, we are upgrading support response standards across all projects under active warranty: for critical incidents, we will respond within 30 minutes and provide a resolution plan and estimated recovery time within 4 hours.

过去我们对不同规模的项目有不同的响应约定，实践中发现这经常让客户在关键时刻需要先做一轮商务沟通才能获得帮助。响应时效不应该取决于合同金额。Previously, response terms varied by project size, and in practice this often meant clients had to go through a round of commercial negotiations before they could get help at a critical moment. Response times should not depend on contract value.

具体包含什么What It Covers

• 7×24 监控告警，故障自动通知到值班工程师24/7 monitoring and alerting, with incidents automatically routed to the on-call engineer
• 严重故障 30 分钟内响应，4 小时内出具处理方案Response within 30 minutes for critical incidents, with a resolution plan within 4 hours
• 每月一次例行巡检并输出巡检报告A monthly health inspection with a written inspection report
• 每季度一次灾难恢复演练并出具演练报告A quarterly disaster recovery drill with a written drill report
• 备份策略按客户要求调整，默认每日全量 + 增量Backup policy tailored to each client, defaulting to daily full plus incremental backups

背后做了什么What Went Into the Upgrade

升级响应标准的前提是值班能力真的跟得上。为此我们在过去一年重做了监控体系：统一告警入口、值班排班系统、以及每个客户独立的健康度看板——任何时候打开，都能看到每个系统在最近 24 小时内的真实状态。Upgrading response standards only works if on-call capacity can actually keep up. To get there, we rebuilt our monitoring over the past year: a single alert entry point, an on-call scheduling system, and a dedicated health dashboard for each client—open it at any time and you see the real status of every system over the past 24 hours.

软件的生命周期比合同长。我们的合同也是。Software outlives its contract. So do ours.