Prologue:
In the world of high-performance computing, where servers hum with the power of many, our latest challenge awaited. Tasked with ethically hacking a $125,000 HPC cluster, we faced four OpenSUSE Linux rack-mounted servers and two Windows-based systems running HyperV. These systems, responsible for running and analyzing molecular simulations, had been handed over to us with only fragmented documentation. Among these remnants were references to user accounts, networking details, a shorthand for a password, and mentions of an RSA key file for SSH access.
Chapter One: Into the Unknown
We began our journey with a strategic choice: rather than using OpenSUSE recovery mode, which might disturb hard-coded uses within applications and services, we opted to investigate the Windows systems and the primary engineer’s old laptop. Armed with administrative passwords for these systems, we had a solid starting point.
Chapter Two: The Labyrinth of Passwords
Using basic system forensics, we discovered the private key file that allowed SSH connections to the Linux servers. Yet, without the correct passwords, our progress was stalled. We turned to the powerful tool Mimikatz on the Windows systems, uncovering several passwords tied to various user accounts. Among these were several that of belonged to the primary engineer and one which featured a sequence familiar from the shorthand notes we had found.
Chapter Three: Unlocking the Secrets
With the engineer’s password, we gained access as a sudo-enabled user but still lacked the ultimate root password needed for full control. By carefully examining the recovered passwords and the shorthand reference, we incremented a known password to gain access to the service account. A search through the bash history file to learn more about how the systems had been configured and used revealed a significant find: a password mistakenly typed into the command terminal. This password, tied to a known vendor and simple in its construction, was the key to our breakthrough.
Chapter Four: Mastering the Cluster
Armed with this new password, we attempted root access and succeeded. Root access to all four Linux servers was now within our grasp, marking the completion of our quest. The HPC cluster, with all its computational prowess, was fully operational under our control.
Epilogue: A Tale Inspired by Middle-earth
The original engineers had drawn inspiration from J.R.R. Tolkien’s epic world, naming their systems after characters from the Lord of the Rings. The four Linux servers were Frodo, Sam, Merry, and Pippin. The NAS VM on one of the Windows hosts was named Bilbo, while other VMs carried the names of various Middle-earth characters. The 10Gb backbone network was whimsically called Hobbiton, and the key file was aptly named “ring,” reflecting the engineers’ vision of their creation as an epic journey and a cohesive system.
Reflections from the Shire
Our adventure through this digital Middle-earth was filled with trials and required both skill and ingenuity. This experience serves as a reminder that even in the realm of cybersecurity, creativity and inspiration can transform a technical challenge into a storied adventure.
Reflecting on our journey, I am reminded of Bilbo Baggins’ words: “It’s a dangerous business, Frodo, going out your door. You step onto the road, and if you don’t keep your feet, there’s no knowing where you might be swept off to.” In cybersecurity, each step can lead to new discoveries and unexpected triumphs.
