Should I self-host Cronitor?
Healthchecks is excellent and installs in half an hour. Cronitor's solo tier is $2/mo. If your time is worth anything at all, the subscription wins — and the monitor watching your cron jobs shouldn't share fate with the box running them.
Independence: self-hosted on the same infra it watches, an outage takes out the alarm too
$24/yr of not thinking
dead-man-switch cron monitoring · alerting (email/webhook/Slack) · status badges · schedules/grace windows
* Percent of the subscription price. Your time is priced at the slider's rate — $0/h ("my homelab time is free and I love it") is a legitimate position. Every number is derived from the record, never hand-written. Price source, checked 2026-08-05.
Runs comfortably on a Raspberry Pi.
The image's built-in healthcheck probes the database connection, and the sqlite default has no writable path in the stock container — it 500s forever until you either mount the sqlite dir correctly or (better) give it the Postgres it actually wants
# Healthchecks — cron job monitoring (tested against Cronitor's job list; see the verdict
# for why the $2/mo subscription out-competes your own attention)
services:
healthchecks:
image: healthchecks/healthchecks:latest
container_name: healthchecks
restart: unless-stopped
depends_on:
hc-db:
condition: service_healthy
environment:
SECRET_KEY: 7f3a9c1e5b8d2f6a0c4e8b2d6f0a4c8e # throwaway test value
SITE_ROOT: http://localhost:8089
ALLOWED_HOSTS: "*"
DEBUG: "False"
DB: postgres
DB_HOST: hc-db
DB_NAME: hc
DB_USER: hc
DB_PASSWORD: hc
ports:
- "8089:8000"
# No healthcheck override: the image ships its own (./fetchstatus.py probes
# /api/v3/status/, which checks the DB connection — sqlite without a writable dir
# fails it, which is why this compose brings postgres).
hc-db:
image: postgres:16-alpine
restart: unless-stopped
environment:
POSTGRES_USER: hc
POSTGRES_PASSWORD: hc
POSTGRES_DB: hc
volumes:
- hc-pg:/var/lib/postgresql/data
healthcheck:
test: ["CMD-SHELL", "pg_isready -U hc"]
interval: 5s
timeout: 5s
retries: 12
volumes:
hc-pg:
Want a guided install instead? caniselfhostit.com/cronitor has AI-agent prompts that assume a bare machine — that's their half of the stool, and it's good.
FULL TIMED LOG
# Timed setup log: Cronitor → healthchecks **Protocol:** v1 · **Verified by:** zernonia · **Date:** 2026-08-10 **Assistant:** claude-code · **Environment:** containerized runner, 2 vCPU class, Docker 29.3 / Compose v5.1 ## Timeline | Step | Time | |---|---| | Read upstream docs, draft compose with healthcheck | ~11 min | | **Boot: all services healthy in 12 s (measured, `compose up --wait`)** | — | | Endpoint-level workflow check (login/health/API serve) + re-run from clean volumes | ~11 min | | **Total: ~35 min** | | Boot and endpoint checks are machine-verified; `setup_min` is the wall-clock total for this session including authoring, diagnosis and re-runs. Endpoint-level ≠ full human UI workflow — dispute anything that doesn't reproduce (CONTRIBUTING.md). ## What broke - The image's built-in healthcheck probes the database connection, and the sqlite default has no writable path in the stock container — it 500s forever until you either mount the sqlite dir correctly or (better) give it the Postgres it actually wants ## Verdict-relevant notes - Same cheap-SaaS pattern as Pinboard: capability is all there; the economics simply never arrive. Run it if you already have a fleet and want internal-only pings.