Uptime: DevFlow's own SLOs and status
Uptime: our own SLOs
We dogfood every part of our product. Our own status page (status.devflow.io) is built on status-pages. Our own pagers fire on the same slo-multi-window-alerting rules we recommend.
our targets
| Service | SLO | Window |
|---|---|---|
| Probe network (writes) | 99.95% successful check execution | 28d |
| Public API | 99.9% availability + p95 < 250ms | 28d |
| Dashboard | 99.5% availability | 28d |
| Webhook delivery | 99.5% delivery within 60s | 28d |
our error-budget posture
We treat the SLO as a hard contract with ourselves. If we burn through the budget mid-window, we freeze non-essential changes (new features) until the window closes.
what we don't promise
- Sub-10s checks are explicitly not an SLO; they're best-effort on the Scale tier. We deliver them most of the time.
- Status-page custom-domain TLS issuance has a 99% target — DNS provider issues sometimes pin us below.
subscribe
Subscribe to status.devflow.io for incident announcements, sub-processor changes (compliance-soc2-gdpr), and monthly attainment reports.
the last 12 months
Probe writes 99.97% / API 99.94% / Dashboard 99.7% / Webhook delivery 99.6%. The full month-by-month breakdown is on the status page.
why we publish this
Our customers' SLOs depend on ours. If you're tracking 99.9% on your service and your monitoring says 99.99% when you're really 99.5%, we've failed you. We'd rather hold ourselves to a bar in public.