type: runbook
status: active
timestamp: 2026-06-22
tags: [runbook, hosting, free-tier, monitoring, uptime, sentry, better-stack, axiom]
status: active
timestamp: 2026-06-22
tags: [runbook, hosting, free-tier, monitoring, uptime, sentry, better-stack, axiom]
Free hosting — monitoring (Better Stack, UptimeRobot, Healthchecks, Sentry, Axiom)
Monitoring free-tier: Better Stack, Sentry, Axiom
Monitoring + uptime + logs — free tiers (2026-06-22)
Primary uptime in 2026 is the custom oriz-status-app (CF Worker + KV cache, self-hosted on Cloudflare Pages). Third-party uptime tools are kept as secondary because no commercial-OK free tier scales to the family’s 45+ endpoint count:
- UptimeRobot became personal/OSS-only in Oct 2024 → unusable for commercial endpoints.
- Better Stack Uptime free is capped at 10 monitors → covers only the top 10 endpoints, not the 45+ fleet.
For errors, Sentry Developer + Axiom Personal still cover the family at $0.
The table
| # | Provider | Free tier | Card@signup | Card to use free | KYC | Verdict |
|---|---|---|---|---|---|---|
| 1 | Better Stack (Uptime) ⭐ | 10 monitors + heartbeats, 1 status page, Slack/email alerts, 100K exceptions/mo, 5K replays | NO | NO | NO | KEEP (commercial OK) |
| 2 | UptimeRobot | 50 monitors, 5-min interval, 1 status page, 3-mo retention. Non-commercial only since Oct 2024 | NO | NO | NO | KEEP (personal/OSS only) |
| 3 | Healthchecks.io ⭐ | 20 checks, 100 log entries/check, 5 SMS credits/mo, unlimited team | NO | NO | NO | KEEP (cron monitoring) |
| 4 | Sentry Developer ⭐ | 5K errors, 50 replays, 5 GB logs, 5 GB app metrics, 5M spans, 1 uptime + 1 cron, 1 user | NO | NO | NO | KEEP |
| 5 | Axiom Personal ⭐ | 500 GB/mo ingest, 25 GB storage, 10 GB-hr query, 30-day retention, 3 datasets, 1 user | NO | NO | NO | KEEP |
| 6 | Better Stack Logs (was Logtail) | 3 GB logs/mo (3-day retention), 30 GB metrics, 3 GB traces, 3 GB warehouse events | NO | NO | NO | KEEP |
How the family uses monitoring
Target end state (in progress, 2026-06):
| Layer | Provider | Cost | Status |
|---|---|---|---|
| Uptime (primary, all 45+ endpoints) | oriz-status-app (custom; CF Worker + KV cache, hosted on CF Pages) | $0 | In build |
| Uptime (secondary, top-10 commercial) | Better Stack | $0 (capped at 10 monitors → cannot cover full fleet) | Active |
| Uptime (tertiary, personal/OSS only) | UptimeRobot | $0 (commercial-use ban since Oct 2024) | Active for OSS sites only |
| Cron monitoring (restic backup, daily catalog rebuild, etc.) | Healthchecks.io | $0 (20 checks) | Active |
| Error tracking (frontend + Workers) | Sentry Developer | $0 (5K errors/mo, plenty of headroom) | Active |
| Log ingestion (Workers logs, Pages function logs) | Axiom Personal | $0 (500 GB/mo is huge) | Active |
Why we’re building oriz-status-app
- Better Stack hard-caps free at 10 monitors. The family runs 45+ endpoints (5 apps + 19 API surfaces + 21 content sites). Better Stack covers only the top 10 → 35+ endpoints uncovered.
- UptimeRobot 2024 ToS change banned commercial use on the free tier. Roughly half the family is commercial-adjacent (paid services, ads-enabled sites), so UptimeRobot is unsafe except on the strictly-OSS sites.
- No third-party free tier scales to 45+ commercial monitors. Conclusion: build our own.
oriz-status-appshape: CF Worker hits each endpoint every 5 min via Cron Triggers, writes result to CF KV (status:<endpoint>:<timestamp>keys with TTL); CF Pages site reads KV and renders the status page. All within Workers Free quota (100K req/day budget; 45 endpoints × 12/hr × 24 = 12,960 checks/day → 13% of cap).
Quirks per provider
- UptimeRobot Oct 2024 ToS change. Free is now “personal, OSS, non-profit” only — commercial use requires paid. Most family sites are personal/OSS so this is fine, but the catalog + paid services (when Razorpay goes live) need Better Stack.
- Better Stack = Better Uptime + Logtail merged. The “Uptime” free tier is 10 monitors; “Logs” tier is separate (3 GB/mo).
- Healthchecks.io is the killer for cron health. Each cron job pings a URL on success; if the ping doesn’t arrive in time, you get alerted. 20 checks covers the family’s restic, mirror cron, daily catalog rebuild, hourly auto-source ingest, etc.
- Sentry Developer. 5K errors/mo + 50 replays + 5M spans is genuinely generous. 1 user limit kills team use; for solo-maintainer family this is fine.
- Axiom Personal 500 GB/mo ingest is the most generous log tier in the table. Comparable to Datadog’s $$$$$ — Axiom does it free.
- Better Stack Logs (ex-Logtail). Merged into Better Stack. 3 GB logs is smaller than Axiom; use Axiom if logs volume matters.
Recommendation for the family
- Uptime (primary): build + run
oriz-status-app(CF Worker + KV + Pages). Only free path to cover 45+ commercial endpoints. - Uptime (top-10 backup): Better Stack — 10 monitors covers the top 10 commercial endpoints as a redundant rail (independent vendor in case our own Worker layer goes down).
- Uptime (OSS-only overflow): UptimeRobot — 50 monitors free, personal/OSS sites only (commercial-use ban).
- Cron job health: Healthchecks.io — 20 checks free.
- Frontend + Worker errors: Sentry Developer — 5K errors/mo free.
- Logs + traces + metrics: Axiom Personal — 500 GB/mo ingest free.
- Replace any of the above with Cloudflare Analytics if the metric fits inside what CF already collects (most “page view + simple latency” metrics do).
Sources
- Better Stack pricing — 10 monitors free
- UptimeRobot pricing + Oct 2024 ToS change
- Healthchecks.io pricing — 20 checks free
- Sentry pricing — Developer plan — 5K errors free
- Axiom pricing — Personal — 500 GB/mo free
- Better Stack Logs (Logtail) pricing