🖭 The “Prestik, Scotch Tape, and Barbed Wire” Data Centre | A Cautionary Tale 🗑️
Encountering makeshift data centres held together with goo, tape & wire
The term "Prestik, Scotch tape, and barbed wire data centre" may sound like hyperbole, but for anyone who has worked in IT long enough, it conjures up vivid imagery of poorly designed, mismanaged facilities held together with makeshift solutions. These data centres epitomize a lack of foresight, investment, and operational discipline, often leading to catastrophic consequences that could have been avoided with proper planning and management.
Symptoms of a "Prestik and Barbed Wire" Data Centre
Limited Documentation and Non-Existent SOPs
Poorly documented systems and the absence of standard operating procedures (SOPs) leave data centre teams flying blind. When an incident occurs, the lack of clear guidelines leads to confusion and delays, exacerbating the problem.
Outcome: Extended downtime, miscommunication during incidents, and a high potential for human error.No Visible Emergency Procedures
Without clearly posted and regularly updated emergency plans, teams lack the ability to respond quickly to critical events such as fires, floods, or power failures.
Outcome: Chaos during emergencies and extended recovery times.Poor Labelling
Inadequate labelling of cables, servers, and equipment results in a spaghetti mess of infrastructure that makes troubleshooting a nightmare.
Outcome: Increased risk of accidental disconnects and delays in resolving issues.Continuous Change
Uncontrolled changes in the environment—without proper change management practices—lead to unpredictable outcomes and heightened risks of failure.
Outcome: A fragile system where any small change can cause cascading failures.Headless Chickens
A lack of coordination between service delivery and operational teams turns incident response into a blame game.
Outcome: Finger-pointing instead of problem-solving, causing delays in service restoration.
Common Problems in Poorly Managed Data Centres
Water Leaks: A Preventable Nightmare
Water is the antithesis of data centres, yet leaks are surprisingly common. These can arise from:
Air handling units icing over or pipe failures.
Kitchens or ablution facilities located above or near the data centre.
External floods seeping into basement data centres.
Why It Matters:
Without flood detection systems and an emergency response plan, water damage can destroy equipment and result in extended outages. If you don’t have a sump pump on hand, you might end up floating out of the data centre in a server cabinet ark.
Prevention:
Implement tested flood detection systems.
Ensure proper drainage and physical barriers.
Conduct routine inspections of water-related systems.
Power Outages | The Achilles Heel
Power continuity is vital, yet poorly managed data centres often fail during outages due to:
Insufficient testing of backup systems.
Generators without proper load-testing protocols.
Inadequate planning for variable speed air handling units.
The Hidden Risk:
A quick five-minute test may not uncover problems that arise during a real outage. For instance, air handling units require time-delayed starts to avoid overloading generators. Controllers managing these systems need to be on UPS to ensure they can manage the staggered restarts.
Prevention:
Conduct full-duration power outage simulations (4–8 hours).
Ensure all critical control systems are connected to UPS.
Time-stagger the start of large power-drawing systems to avoid generator overload.
Monitoring | The Missing Link
Many data centres fail because they neglect real-time monitoring. IT teams rely on network monitoring tools, while facilities staff rely on clipboards. This disconnect can result in unnoticed failures.
Real-Life Example:
A data centre's generator stopped charging the UPS batteries, but no one noticed because the UPS didn’t have a web or Ethernet module. The result? A complete power failure that could have been avoided with proper monitoring.
Prevention:
Equip UPS systems with web or Ethernet modules for real-time monitoring.
Integrate facility monitoring into the IT network for unified visibility.
Conduct regular cross-functional audits to ensure all systems are monitored effectively.
The Cost of Neglect
When management prioritizes budget constraints over operational excellence, the consequences are severe:
Major Incidents: Unplanned outages result in productivity loss and financial damage.
Brand Reputation: Customers and stakeholders lose trust in your business.
Inefficiency: Time spent patching systems together detracts from innovation and growth.
What’s the Solution?
Invest in Infrastructure: Build with quality materials and systems designed for data centres, not office spaces.
Implement Rigorous Testing: Test backup systems under realistic conditions to uncover weaknesses before they become disasters.
Bridge the IT-Facilities Gap: Foster collaboration between service delivery and operations teams.
Adopt Proactive Maintenance: Regularly inspect and maintain equipment to address issues before they escalate.
Plan for Emergencies: Have well-documented and visible emergency procedures and ensure staff are trained to execute them.
Wrap
A “Prestik, Scotch tape, and barbed wire” approach to data centre management is a recipe for disaster. While it may save money in the short term, the long-term costs of outages, lost productivity, and reputation damage far outweigh the investment in building and maintaining a robust, well-managed facility. By addressing the symptoms of poor design and embracing proactive strategies, businesses can ensure their data centres deliver the reliability and performance that modern businesses demand.
