The Great AWS Hardware Incident of 2025: A Comedy in Three Acts

Act 1: The Midnight Panic

Thursday night, 12:57 AM EST (or as I call it, “why am I still awake” o’clock)

Nothing ruins a late night quite like seeing an AWS health alert in your inbox. Well, that and a Cloudflare Error 521 page where your website used to be.

“We are investigating an issue that impacts the availability of some EC2 instances…”

Translation: “Your server went to hardware heaven, hope you had backups lol”

Act 2: The Investigation

Me: frantically SSHs into server

Server: “Welcome to Bitnami! Everything is fine! 😊”

Also Server: serves the Apache default page instead of the actual website

Me: “But… you’re running?”

Server: “Oh yeah, totally running. Just not running YOUR stuff.”

Turns out, I had TWO Apache installations living their best lives on my EC2 instance:

• • System Apache (Debian): The overachiever who woke up first after the restart, grabbed port 80, and started serving… the default Apache page. Classic.

• • Bitnami Apache: The one with my actual WordPress site, sitting in the corner like “bro, port 80 is taken, what do you want me to do?”

Act 3: The Resolution

The fix? A tale as old as time:

bash

# Stop the impostor
sudo systemctl stop apache2
sudo systemctl disable apache2

# Start the real MVP
sudo /opt/bitnami/ctlscript.sh start

Plot twist: The Elastic IP I set up months ago saved me from DNS hell. Past me was clearly smarter than present me.

Lessons Learned

1. 1. AWS hardware failures happen – They’re like that friend who shows up to help you move but brings the wrong truck

1. 1. Always have an Elastic IP – Because updating DNS records during a crisis is nobody’s idea of fun

1. 1. Two Apache installations = One headache – There can be only one (on port 80)

1. 1. Bitnami vs System packages – Choose your fighter wisely

The Moral of the Story

Having your website down is never fun, but at least it makes for a good story. Additionally, consider setting up Auto Recovery and some monitoring next time.

Status: Website restored
Downtime: ~40 minutes
Coffee consumed: Too much

PS: If you’re reading this, my site is back up. If you’re not reading this… well, we’ve got bigger problems.