This post was originally posted on Coded Geekery, my personal blog about finding a work-life balance, but I wanted to share it here as well since it's about the Flybase status page.
We built our own status page for Flybase, you can see it here.
It was built as a static HTML page on Github Pages. It uses Github Issues to report any incidents and Uptime Robot to monitor our sites.
This idea is based loosely on the statuspage repo created by @pyupio, but simplified, as I wanted this to be pretty much automated, plus we already use Uptime Robot for monitoring, so combining Uptime Robot with Github Issues works great.
To get started, you'll want two things:
- An Uptime Robot account
- A GitHub repo where you can throw your site up and use the issues system.
Create a branch in your repo called
gh-pages, this is where your files will sit.
Ok, let's build our status page:
1. create index.html
First, create our
2. Create script.js
Next, we'll create
script.js, this is the file that talks to our services.
Replace the following variables with actual lines:
YOUR-UPTIME-ROBOT-API-KEY-2: Uptime Robot's default API key is universal, but is also read and write. We want this monitor to be read-only so we have to create an API Key for each site we are creating. This is an array of keys.
To add more sites, just add a new line and add a new monitor key.
YOUR-GITHUB-USERNAME: Your Github username where the repo you created lives.
YOUR-GITHUB-REPO: The Github repo you created to use.
2. Create style.css
Finally, we want to create our
style.css file. Just copy this entire block into the file.
4. Git it Going
Once you've created your files, you want to put them on your repo, you can either create them directly from the github.com interface, or you can create them locally and commit them.
If you want to add this to a specific domain, then create a file called
CNAME and store your domain or subdomain in there.
Finally, create a file called
.nojekyll which tells Github Pages that this is a strictly static site.
To customize Github Issues, we set up labels to identify issues:
operationalmeans all systems good.
investigatingmeans under investigation.
outageto identify an outage.
degradedto identify an issue causing degraded performance.
On top of that, you can add labels that start with
system: and they will show what system the issue is related to. For example
system:blog would show an issue with our blog.
Labeling an issue with any of these tags will reflect on the status page.
This status page works pretty well, and was useful last week with the AWS outage that happened. It showed the status of our various services, and let us push updates via Github Issues that showed up below.
I do plan on making an update at some point to take into account comments inside issues.
This is a basic status page, but it helps show people what is happening with your sites, and keep everything nice and transparent.
Originally posted on Coded Geekery, my personal blog about finding a work-life balance