Ping

A simple status page built over Cloudflare Workers, D1 and Vercel (Next.js 13).

Since this project started as a way to test new tech, it is not very tidy or properly split into components, but it works nicely!

It has some nice little features:

Features

Host dependent

For single developers and startups, only deploy once! You can configure multiple pages depending on host and assign monitors to each

Extremely fast, built on Cloudflare Workers, Vercel + React server components, and Cloudflare D1.
Configurable. Monitors accept special configurations to decide how to run them (only http for now) and when to create automatic incidents
Schedulable maintenance pages

Planned features (no date set)

Heartbeat monitors
Configuration page (OAuth & Basic HTTP auth)

Screenshots

Landing page

Maintenance page

Incident history

Incident updates

Why does this project exist ?

I wanted to test some new tech, and ended up deciding on a status monitor, because it allowed me to test them all at the same time:

Next13 + Server Components
Cloudflare D1
NodeJS's own test runner
Turborepo

Understanding apps & components

This repo is divided in 3 different components

pinger: Is what does the actual monitoring, reads/writes to Cloudflare D1
front: Is your status page written on Next.js 13
api: Is the management API and used by front-end

There is also a package called db. This package is shared between pinger and api to read/write from the database. It is a small wrapper over D1's API to simulate an ORM.

Embeds

The app also allows you to call any "page" via the /embed link

Ex: https://example.com/embed/my-page

These are special pages where the background is transparent and there is no page title nor menu. They allow you to embed your monitors/performance trackers in different pages.

Here's an example in Notion:

Back of the envelope calculation for disk size

* 1 monitor ~= 250B
* 1 check ~= 215B (max 1min)

Average config:
* 6/7 products & internal stuff
* 3/4 monitors per product
* => 28 monitors total

@ 1 check/min
1y of checks = 60 * 24 * 365 = 525600 checks
=> 113 MB for 1 monitor
=> 3GB for 28 monitors

@ 1 check/10min
6 checks/h
144 checks/day
52560 cehcks/year
=> 11 MB for 1 monitor
=> 308MB for 28 monitors

@ 1 check / 30min
2 chekcs / h
48 checks / day
17 520 checks/year
=> 4.38 MB for 1 monitor
=> 122.64 MB for 28 monitors

Requirements
4GB disk space @ 1 check / min
500MB disk space @ 1 check / 10min
150 MB disk space @ 1 check/ 30min

As of today (2023-06-11) D1 only supports 100MB of data.

jsmrcaga/ping