一人工程 · solus opus

← 全部作品

一人工程uptimesolomonitoring

solo engineer 没有 uptime,我自己就是 uptime

团队有 uptime。

Uptime:在线时间。Availability:99.9% / 99.99% / 99.999%。Uptime monitoring:Datadog / Pingdom / StatusGator。Datadog:APM 服务。Pingdom:uptime 检查。Status page:status.example.com。On-call rotation:uptime 挂了谁修。

每个 startup 1 个 monitoring tool $500/月 + 1 个 on-call rotation。

solo engineer 没有 uptime。

我自己就是 uptime。

我 commit log = uptime。我 Vercel 自动 deploy = uptime monitor。我不需要 Datadog、不需要 Pingdom、不需要 status page、不需要 on-call rotation。

我没 uptime 的必要——访问量 1000,挂了 = 没人发现。我没 availability 的必要——没 SLA 承诺。我没 99.9% 的必要——没 KPI。我没 Datadog 的必要——访问量不够大。我没 Pingdom 的必要——我自己看 Vercel。我没 status page 的必要——没 enterprise customer。

团队的 uptime 价值是「让 customer 知道服务在线 + 快速恢复」。前提是有 customer、需要 transparency、需要 commitment。solo engineer 没这个问题——没 enterprise customer、不需要 transparency、commit log = commitment。

一人工程的 uptime = Vercel 部署状态。不是 Datadog $500/月。