Quick start

Tickwatch does not connect to your servers and runs nothing on them. It works the other way round: the job reports that it finished. No report in time — we write to you.

Which gives the main convenience: no open ports, no public IP, no agent. Outgoing HTTPS from the machine is enough.

Three steps

• Create a monitor and set the schedule — the same one your crontab already has.
• Copy the ping address from the monitor page and append it to the job.
• Wait for a run, or press “Test ping” in the panel to confirm the signal gets through.

/path/to/job.sh && curl -fsS https://ping.tickwatch.dev/YOUR-KEY

The && matters more than it looks: the ping is sent only on a zero exit code. With a semicolon the monitor stays green even after the job fails — and you would learn about the problem exactly when you did before: from a customer.

What next

The basic setup catches only a missing job. Add start and failure signals, and you will also learn that a job crashed or hung — with the tail of its output included in the alert.

#!/usr/bin/env bash
set -euo pipefail

PING="https://ping.tickwatch.dev/YOUR-KEY"

# Сообщаем о начале: с этого момента считается длительность.
curl -fsS -m 10 --retry 3 "$PING/start" > /dev/null

if out=$(/path/to/job.sh 2>&1); then
  curl -fsS -m 10 --retry 3 --data-raw "$out" "$PING" > /dev/null
else
  code=$?
  # Хвост вывода уедет в письмо — разбираться можно, не заходя на сервер.
  curl -fsS -m 10 --retry 3 --data-raw "$out" "$PING/$code" > /dev/null
  exit $code
fi