October 22, 2026

When Server Rooms Go Underground

Last week I read that Russia is considering putting data centres in abandoned coal mines, roughly 300 metres underground, partly to protect infrastructure from Ukrainian drone attacks and partly for cooling and security benefits. Yes, really. Source: TechRadar, 'Data centers can be located in exhausted open pits': Russia eyes 300-meter-deep coal mines to shield server infrastructure from Ukrainian drones.

On one level, it sounds absurd. On another, it is the logical endpoint of something most businesses still underestimate: infrastructure is no longer just an IT decision. It is a resilience decision. A security decision. A commercial decision. Sometimes, frankly, a survival decision.

And that is where this gets interesting for founders, CTOs, CIOs and anyone hiring into infrastructure, cloud, cyber and data roles.

Infrastructure has entered its paranoia era

For years, plenty of leaders treated infrastructure like plumbing. Necessary, expensive, a bit boring, best hidden behind a cupboard door while everyone got excited about product, growth and AI slides.

That era is over.

Now infrastructure sits in the middle of a messy Venn diagram of cyber risk, physical risk, sovereignty, uptime, energy cost and operational resilience. You are not just choosing where to host workloads. You are choosing your exposure.

The Russia coal mine idea is an extreme example, but it highlights a broader truth. When the environment gets hostile, businesses start redesigning systems around threat models, not convenience.

That applies well beyond military conflict.

Think about the pressure points already shaping decisions in the UK and Europe:

• Energy cost volatility
• Grid constraints
• Climate and cooling concerns
• Physical site vulnerability
• Data sovereignty and regulation
• Cyber attacks on critical infrastructure
• Supply chain fragility

If your infrastructure plan still reads like it was written in 2019, with a bit of cloud sprinkled on top like parsley, you may have a problem.

The weird idea is not the point

Let us be honest. Most organisations are not about to move compute into a former coal mine. If your ops strategy currently includes 'find cave', we should probably talk.

But dismissing the story because it sounds dramatic misses the lesson.

The real point is this: serious organisations redesign infrastructure when the cost of fragility becomes impossible to ignore.

That redesign might mean:

• Repatriating some workloads from public cloud
• Splitting critical systems across regions
• Building true disaster recovery, not the PowerPoint version
• Investing in edge or sovereign infrastructure
• Hardening physical and cyber controls together
• Hiring leaders who understand resilience at system level

In calmer markets, these conversations often get delayed because they are not shiny. They do not demo well. No one gets a standing ovation for saying, 'Good news, we have reduced concentration risk.'

Until something breaks.

Then suddenly the boring people become very interesting.

Most companies do not have an infrastructure problem. They have a leadership problem.

Here is the part I care about most as a recruiter.

When companies struggle with infrastructure, cloud migration, cyber posture or data platform resilience, the issue is often framed as a tooling gap. Wrong vendor. Wrong stack. Wrong architecture.

Sometimes that is true.

Often, though, the deeper problem is that nobody in the room truly owns the outcome with the right blend of technical depth, commercial judgement and backbone.

You need leaders who can answer questions like:

• What are our real single points of failure?
• What would 24 hours of downtime actually cost us?
• Which systems are mission critical, and which are just loud?
• Where are we overexposed geographically, commercially or operationally?
• Have we built resilience into the architecture, or merely talked about it in risk meetings?

This is where hiring gets serious.

A strong Head of Infrastructure, Cloud Director, Platform Lead or CISO does not just keep the lights on. They make sure the building is not made of papier-mâché in the first place.

Too many businesses hire these roles too late, define them too narrowly, or expect them to perform miracles with a budget that looks like a rounding error.

Then everyone acts shocked when the transformation stalls.

Resilience is now a hiring strategy

This is the shift I think more leaders need to clock.

Resilience is not only designed through technology choices. It is built through hiring choices.

If you are scaling, modernising, or operating in a high-risk environment, your recruitment strategy should reflect that reality.

Ask yourself:

• Do we have infrastructure leadership that scales with our risk profile?
• Are our cloud and platform hires strategic thinkers or just safe pairs of hands?
• Have we hired for yesterday's estate or tomorrow's threat landscape?
• Can our technology leaders translate resilience into business language the board understands?

That last point matters more than people think.

The best infrastructure and cyber leaders are interpreters. They can talk latency with engineers, risk with the board, budgets with finance and trade-offs with product. They know when to push for gold-plating and when to say, 'No, this will do.'

That is rare talent. And it is exactly why hiring by keyword matching is such a waste of everyone’s time.

The practical framework: pressure-test your setup

You do not need a decommissioned mine to think more clearly about resilience. You do need honesty.

Here is a simple framework I would use internally.

Map the blast radius

Identify what happens if a critical system, site, provider or team goes down. Be specific. Not vague. Not optimistic. Specific.

Questions to ask:

• What breaks first?
• What breaks next?
• Who notices?
• How long until customers feel it?
• How long until revenue feels it?

Audit concentration risk

Look at where you are over-reliant on one region, one vendor, one data centre model, one team or one individual who 'knows how it all works'. That last one is particularly dangerous. Every company has a legend in a hoodie. Legends go on holiday.

Test leadership depth

Do you have people who can lead under stress, make trade-offs fast and communicate clearly when things go sideways? Capability under pressure is not visible on a tidy CV. It comes out in how people have handled real complexity before.

Hire before the pain peaks

If you wait until you are mid-crisis to hire infrastructure or cyber leadership, you are shopping for lifeboats after the ship has hit the iceberg. The best people are rarely available at panic speed.

What smart leaders should do next

If this story triggered a mild eye-roll, fair enough. If it triggered a more serious thought about your own infrastructure exposure, even better.

Here is the actionable bit.

• Review your infrastructure strategy through a resilience lens, not just a cost lens
• Stress-test your operational dependencies, including physical and geopolitical ones
• Reassess senior hiring plans for infrastructure, cloud and cyber
• Make sure your technical leaders can communicate board-level risk clearly
• Stop treating resilience as a back-office concern. It is now part of competitive advantage

The companies that navigate uncertainty best will not necessarily be the ones with the flashiest tech stack. They will be the ones with the clearest thinking, the strongest operators and the courage to invest before the market forces their hand.

Conclusion

Server farms in coal mines make for a cracking headline. But the headline is not really about mines. It is about mindset.

When the world gets less predictable, infrastructure stops being invisible. It becomes strategy in steel-toe boots.

And if you are still hiring these roles like they are glorified caretakers of legacy kit, you are missing the plot entirely.

The future belongs to businesses that treat resilience as a boardroom issue, not a basement issue. Preferably before they feel the need to move the basement 300 metres underground.

Back to news