The smell of turkey legs and the sound of lutes. The taste of panic and the weight of a BlackBerry that held my company’s life in its electronic pulse.
What still galls about this story is how ordinary the cause was. Every protection in that building had been bought, installed and signed off, and they failed on the same morning because they shared the same maintenance cycles and oversight gaps, the kind I’d bet still sit in most computer rooms today. The consultant’s rebuild makes a great ending, and no company should ever have to count on one.
The BlackBerry buzzed at 7:03 AM, cutting through the morning stillness like a blade.
The consultant’s rebuild makes a great ending, and no company should ever have to count on one.
I was already in my Saturn, driving toward Fresno with camera gear rattling in the back: strobes, reflectors, the kind of equipment that promised a perfect day of Renaissance Faire photography. Three gorgeous models waited ahead, elaborate costumes and morning light that would be magical.

The three ladies waiting for me at the Kearney Park faire in Fresno.
The caller ID glowed: Operations Manager.
“What do you want me to do about the computer crash?” His voice carried an edge sharp enough to cut glass.
I pulled over, hazard lights painting the asphalt amber. Cars whooshed past, their drivers heading to normal jobs where computers either worked or they didn’t, where a disaster meant spilled coffee.
“What crash?”
The Cascade Begins
“Walk me through it,” I said, gripping the steering wheel like it could anchor me to a world that still made sense.
“Power went out around 3 AM. Generator didn’t start. UPS held for maybe twenty minutes, then one of the redundant units failed.”
Through my windshield, California morning light painted the world gold and innocent. Behind me, my company was hemorrhaging digital blood.
“That’s when we found out the disk cabinets were cross-plugged. And Richard…” His voice trailed off like a man afraid to speak his own death sentence.
“Tell me.”
“The disk controller battery was dead. The SAN table corrupted when everything went down hard.”
I shifted the BlackBerry to my other shoulder and waited for the killing blow.
“Backup tapes failed. Disaster recovery site is down too.”
Every safety net we’d built. Every redundancy we’d paid for. Every “this will never happen” scenario. All of it had collapsed like a house of cards in a hurricane.
“How bad?”
“We have no way to recover anything. Five hundred disks, scrambled like electronic eggs. The company’s dead, Richard.”
Medieval music drifted across the parking lot, sweet and oblivious. Somewhere, a blacksmith’s hammer rang against anvil.
Managing the Apocalypse
I made a choice that felt counterintuitive but was probably the only rational one: put the Saturn in drive and head toward the Faire.
What was I going to do in person that I couldn’t do remotely? Stand in the data center and watch the expert work? The consultant was already on-site doing everything that could be done. Driving back would accomplish nothing except putting me in a room where I’d be useless and anxious.
I’d make the same call today. A manager who rushes back to hover over the one person who can fix the problem slows that person down, and the business needed me on the phone with the people keeping it alive.
The BlackBerry became my command center. Everyone was calling. The boss, voice cracking with barely controlled panic. Sarah, our operations director, a woman who could organize chaos into profit. Accounting, already calculating severance packages. Warehouse managers staring at silent screens that yesterday had orchestrated the movement of millions in inventory.
Sarah became my lifeline. While I drove through suburbs that had no idea my world was ending, she was transforming a digital company back into something that ran on paper and people.
“We’re going analog,” she told me. “Paper forms, hand-written orders, actual conversations instead of emails. If the computers are dead, we’ll bring this place back to life with pens and stubbornness.”
Meanwhile, I was supposed to capture beauty. Three models in Renaissance finery: velvet and silk that cost more than most people’s cars, jewelry that caught light like trapped fire. Canceling wasn’t an option. These women had cleared their schedules, driven hours, spent small fortunes on costumes.
So I found myself directing a strange kind of theater: “Turn toward the light, perfect, hold that sword higher.” BlackBerry buzzes. “Richard here. How many orders did you hand-process? Really? That many?” Camera clicks. “Beautiful! Now let’s try something fierce, like you’re ready for battle.”
Because in a way, that’s exactly what we were all doing.
The Professional
My consultant was ex-Special Forces, a Fallujah veteran. I hadn’t hired him for his war stories. I’d hired him because he was the kind of person who looked at impossible situations and saw puzzles to solve instead of reasons to panic.
While everyone else ricocheted between denial and despair, he just worked. No drama, no declarations. He treated our company’s death throes like an especially interesting technical challenge.
For hours, he methodically tried every conventional approach to resurrection. Standard backup restoration failed. Disaster recovery procedures failed. Backup table reconstruction failed. Low-level disk analysis failed.
Every one of those procedures had been sold as the thing that would save the day, and each one assumed some other piece of the system was still standing. I think that hidden assumption sits inside most recovery plans written today, and nobody finds it until the morning everything fails together.
The BlackBerry updates grew shorter and grimmer: “Still working.” “Trying something else.” “No luck.”
Then at hour seven, the call that changed everything:
“Richard, he says he has an idea.”
I was coaching a model through a fierce battle pose, sword raised like she was defending the last castle on earth.

Battle poses between phone calls, Kearney Park, Fresno.
“What kind of idea?”
“He says you don’t want to know.”
When someone whose expertise was forged in urban warfare tells you that ignorance might be bliss, you’re about to witness either a miracle or a war crime. Maybe both.
“Should he proceed?”
“Um, yeah.”
Forty-five minutes of silence followed. The longest forty-five minutes in the history of measured time. No updates, no status calls, nothing but the sound of my heart hammering against my ribs while lutes played and children laughed and my entire professional existence hung by a thread I couldn’t see.
Then it rang.
“We’re up and running.”
I sat down hard on a hay bale, still holding my camera, surrounded by people in crowns and chainmail who had no idea they’d just witnessed a resurrection.
From a 3 AM power cut to full recovery: the hours the company spent with no working systems.
Later, I asked him what magic he’d performed. He explained it like reading ingredients off a cereal box: he’d rebuilt the SAN allocation tables from memory. The map that told our storage system where every piece of data lived across five hundred disks. He’d recreated it from scratch, trusting his recollection of the original installation against the certainty of total annihilation.
“Wasn’t that incredibly risky?”
He shrugged. “Everything was already gone. Risk is relative.”
My take
Hire the person who can rebuild the map from memory, and then make sure nobody ever has to. Our SAN allocation tables were gone, and one engineer recreated them from his recollection of the original install while I sat on a hay bale holding a camera. It worked. That was a miracle, and you can’t schedule miracles. Document the systems that hold everything together, keep copies somewhere else, and keep that engineer happy.
How did the company survive the data center disaster?
| Stage | What happened | Lesson |
|---|---|---|
| Power and generator | Power went out around 3 AM and the generator didn’t start. | Assume systems will crash when you can least afford it. |
| UPS | It held for maybe twenty minutes, then a redundant unit failed. | Redundancy isn’t enough if backups share common failure points. |
| Disk storage | Cabinets were cross-plugged and a dead controller battery let the SAN table corrupt. | Shared maintenance cycles and oversight gaps let protections fail together. |
| Backups and DR | Backup tapes failed and the disaster recovery site was down too. | Test disaster recovery under realistic conditions, including cascade effects. |
| Operations | The operations director moved the business to paper forms and hand-written orders. | The human element is both your greatest vulnerability and your salvation. |
| Recovery | The consultant rebuilt the SAN allocation tables from memory, and every bit of data came back. | Document critical systems and keep copies elsewhere; you can’t schedule miracles. |
He saved us. Every bit of data was recovered. The company survived. Sarah’s team had processed more orders manually that day than we’d processed with computers the day before.
That disaster taught me lessons I’ve carried for decades. Redundancy isn’t enough if all your backups share common failure points. Our tape, disk, and DR site failures were all connected to the same maintenance cycles and oversight gaps. Test your disaster recovery under realistic conditions, because simulated failures miss the cascade effects that make real disasters devastating. And the human element is both your greatest vulnerability and your ultimate salvation. Someone should have caught the failing tape backups. But when everything else failed, human expertise and memory saved us.
I still have those photographs from that day.
They’re among my best work: models transformed into medieval warriors, light captured like liquid gold. But every time I look at them, I see the weight of a BlackBerry against my shoulder. I smell roasted almonds mixed with the metallic taste of panic. I hear Sarah’s voice, steady as stone, rebuilding an empire with paper and stubbornness.
Sometimes the most important systems are the people who refuse to accept that impossible situations are impossible.
Frequently Asked Questions
Redundancy isn’t enough if all your backups share common failure points.Share on X
