Key takeaways

  • Try remote management first — IPMI, iLO, iDRAC, KVM-over-IP, or your colo's NOC.
  • Call a smart hands provider if remote options fail — they'll dispatch an engineer to your rack, typically within 2 hours in London.
  • Have cabinet numbers, access codes, and a runbook ready so the engineer can act fast.

Your server just went down. It's 2am. You're not in London. And the data centre won't let a random contractor past security.

This is the guide you need right now — step by step, what to do when your server crashes at a UK data centre and you can't (or don't want to) drive there yourself.

Step 1: Try Remote Access First

Before you send anyone to site, exhaust your remote options:

  • IPMI / iLO / iDRAC — try a remote power cycle through the out-of-band management interface
  • KVM-over-IP — if your colo provides it, connect and check the console
  • PDU power cycle — some smart PDUs let you remote-cycle individual outlets
  • Ask your colo's NOC — some facilities offer basic remote hands (power cycle a server) included in your contract

If none of that works, or you don't have remote management set up (lesson learned), you need someone physically at the rack.

Step 2: Call a Smart Hands Provider

This is what "smart hands" means — you call a provider, describe the problem, and they send a certified engineer to your data centre to fix it.

What a smart hands engineer can do for you:

  • Reboot your server (physical power cycle)
  • Check for blinking LEDs, error codes, or alarms
  • Swap a failed hard drive, PSU, or memory module
  • Re-seat cables and network connections
  • Run diagnostics and report back what they see
  • Photograph the rack so you know what's happening
DACPROS provides smart hands 24/7/365 across the UK

Call 0333 444 8301 and we'll dispatch an engineer — typically within 2 hours for London facilities like Telehouse, Equinix, and Global Switch.

Step 3: Know Which Data Centre You're In

This sounds obvious, but when you're panicking at 2am it matters. The engineer needs:

  • The data centre name and building (e.g. "Telehouse North, Docklands")
  • Your suite, cage or cabinet number
  • Access credentials or authorised contact details
  • A description of the problem and what you've tried

Keep this info in a runbook. If you don't have one, make one after this incident.

Step 4: What to Expect When the Engineer Arrives

A good smart hands provider will:

  1. Check in with the data centre's security and access control
  2. Locate your cabinet or cage
  3. Visually inspect the equipment — LEDs, airflow, cable state
  4. Carry out the work you've requested (reboot, swap, diagnose)
  5. Photograph the before and after state
  6. Report back to you with findings and next steps

At DACPROS, every visit is documented with photos, a written report, and a timestamped log.

Step 5: Prevent the Next 2am Panic

Once you're back up, do these things:

  • Set up out-of-band management (IPMI/iLO/iDRAC) on every server
  • Get a smart PDU with remote power cycling
  • Put a smart hands provider on retainer so you're not Googling at 2am
  • Write a runbook with cabinet numbers, access codes, escalation paths
  • Label everything — if someone else has to work your rack, they need to know what's what

Who to Call

If you need someone at your data centre right now:

  • DACPROS — 0333 444 8301 — 24/7/365, London hub, UK-wide coverage, badged at Telehouse, Equinix, Global Switch and more
  • Your colocation provider's NOC — for basic remote hands (power cycles)
  • Your hardware vendor's support line — for warranty replacements

Bottom Line

A crashed server doesn't have to mean a 3-hour round trip at 2am. Smart hands exists precisely for this — certified engineers who are already near the data centre, already badged for access, and available around the clock.

If this is the first time you've Googled "my server is down" — you probably need a smart hands provider on speed dial. Call us on 0333 444 8301 or email .