Solo founder on-call without PagerDuty
PagerDuty and Opsgenie assume rotations, escalation policies, and managers who approve on-call pay. A solo founder at 2 a.m. needs a phone buzz, a rollback checklist, and maybe one status post — not a schedule swap with themselves.
StillOnline covers the lightweight layer: HTTP checks on your production URL, Telegram owner alerts via the official StillOnline bot, and an optional public status page so support macros stay honest. Free is one channel; Pro adds the rest when revenue justifies it.
Quick answer
Skip PagerDuty until more than one human shares production duty on a calendar. StillOnline sends owner alerts to email, Telegram (StillOnline bot), or Slack after failed probes — Free picks one channel. Add a hosted status page at stillonline.tech/.../s/... when buyers ask “is it down?” (Telegram setup). Escalate to full incident tooling when missed pages cost money or sleep.
Stack that fits one person
| Layer | Solo-friendly tool | When to outgrow |
|---|---|---|
| Detection | StillOnline HTTP probe | Multi-region SRE team |
| Alert | Connect Telegram → StillOnline bot | Need phone tree + manager |
| Comms | StillOnline public page + manual post | Dedicated comms lead |
| Postmortem | Notion doc | Compliance-driven ITSM |
PagerDuty shines when two pagers must fire in sequence. With one phone, Telegram is faster to configure (email alerts if you hate mobile apps).
Five-minute on-call bootstrap
- Ship
GET /health(quickstart). - StillOnline project on production URL.
- Settings → Connect Telegram → Start the StillOnline bot (guide).
- Snippet in personal runbook: mute alert only after green probes return.
- Paste status URL in docs when support volume appears (launch checklist).
Free limit: one owner channel — Telegram or email or Slack. Pick the device that is always on you.
When to add PagerDuty (or similar)
- Two+ engineers with fair rotation requirements.
- Contractual response time measured in minutes with fines.
- Integrations with enterprise ITSM (ServiceNow, Jira Service Management).
- You already pay for a suite that bundles paging.
Until then, enterprise status stacks are usually heavier than the problem.
Night-of playbook (keep it short)
- Telegram ping → open StillOnline dashboard.
- Confirm probe URL — not CDN blip (redirects/antibot).
- Roll back or toggle maintenance (scheduled maintenance).
- Post incident on status page (template).
- Reply to support with the same link.
Related guides
- Telegram owner alerts
- Email uptime alerts
- Side project production checklist
- When you don't need a status page
FAQ
Does StillOnline replace PagerDuty entirely?
For solo on-call, often yes for detection + notify. It does not provide on-call schedules, escalation tiers, or phone bridges. Add PagerDuty when a second engineer must be woken automatically.
Should solo founders enable public StillOnline pages?
When external users exist — yes for trust. Pre-revenue, alerts-only is fine (when you don't need a page).
How do I avoid alert fatigue on StillOnline?
Point checks at stable /health endpoints, not login pages that redirect (probe pitfalls). Tune maintenance windows instead of deleting checks.