Hardware

Datacenter hardware destruction: decommissioning all equipment with proof of data destruction

4 July 2026 · 12 min read · DeSnipperaar

Decommissioning a server room or datacenter is more than hauling away servers. A rack holds storage arrays, switches, routers, firewalls, backup tapes and loose drives, all with data or configuration on them. Datacenter hardware destruction means you have every one of those components irreversibly ground up and checked off by serial number. This article describes the complete datacenter disposal: the data risk per component, the inventory, the choice between pull-and-shred and whole-unit, the cost and the conclusive certificate you end up holding.

A move, a cloud migration or a hardware refresh often frees up a whole server room. Attention then goes to the servers, while the risk is spread across the entire setup. A discarded firewall, a forgotten tape or a loose drive in a drawer holds sensitive data just as much. Datacenter hardware destruction tackles the risk in one go, from the first drive to the last network component.

What does datacenter decommissioning mean?

Server room decommissioning is the controlled retirement of a complete IT setup. It starts logically, taking systems out of the cluster and bringing storage offline, and ends physically with the destruction of every carrier. Datacenter disposal is broader than server-only work, because it also covers network gear, tapes, loose media and the racks themselves. The aim is simple: no carrier may leave the site while the data on it is still readable. The server-specific deep dive is in server destruction at end-of-life.

The data risk in every component

The misconception is that only drives hold data. In a datacenter, sensitive information lives in almost every component.

• Servers. Several HDDs or SSDs, often in RAID, plus NVRAM and RAID cache.
• Storage arrays and SANs. Dozens to hundreds of drives per cabinet, the most data-intensive part.
• Switches, routers and firewalls. Configurations, VPN keys and certificates in flash memory.
• Backup tapes. Full backups, sometimes years old, in the tape library.
• Loose drives and SSDs. Spare units and faulty carriers in drawers.
• Racks and PDUs. No data, but a waste stream for recycling.

Anyone who destroys only the servers leaves the firewall with all the access rules and the tapes with old backups out of scope. Data destruction is only complete when every component goes into the same chain.

Start with an inventory and asset register

The basis of every decommissioning is an accurate count. Record per device the asset tag or serial number, the type and number of internal carriers, any encryption present and the classification of the data. This asset register is later the basis of the certificate. Reckon on roughly a working day of inventory for a filled 42U rack. A good list prevents a carrier going unexplained afterwards, and the bigger the setup the more this step matters. For the technical context of rack setups see servers and racks at end-of-life.

Pull-and-shred or whole-unit?

There are two approaches to in-rack data sanitization. With pull-and-shred we remove only the carriers, register them by serial number and grind them up, and the empty chassis goes back to the supplier or to recycling. This is the usual choice for a lease return. With whole-unit the complete unit goes through a heavy shredder, so no forgotten memory module or cache card is left behind. If in doubt, choose whole-unit or have the cache modules taken along separately. The sensitive data is almost always on the drives and SSDs, but also think of NVRAM and removable flash modules.

Which DIN level do you need?

The DIN 66399 standard describes the H-classes for data carriers, where H stands for hard drives. The higher the number, the smaller the particles and the higher the certainty.

Level	Maximum particle size	Suitable for
H-3	Coarser particles	Ordinary business data, low risk
H-4	Small particles	Personal data, the workable minimum
H-5	Very small particles	SSDs, social security numbers, medical and special data

For classic server hard drives H-4 is usually enough. If you work with special personal data or the arrays contain SSDs, choose H-5, because the small memory chips have to be ground more finely. More on this is in DIN 66399 explained.

Billing per item by serial number

An honest provider does not bill by the kilo or by the hour, but per item by serial number. You know in advance exactly what you pay, and you get conclusive proof, because every serial number in the quote comes back on the certificate. In a datacenter that is extra important, because a storage array soon counts a hundred drives. So ask for a price per carrier and per unit, with serial numbers, which keeps you in control of exactly what was taken and destroyed.

What determines the cost?

The price of a datacenter disposal depends on a few factors. Knowing them lets you compare quotes properly.

• The number of carriers and units. You pay per item. At larger numbers the price per carrier falls.
• Whole unit or only carriers. Whole-unit requires a heavier machine than loose drives.
• The level. H-5 requires finer grinding than H-4.
• The distance. Within 20 km of Amsterdam we charge no call-out fees.
• The mix of streams. Servers, tapes, switches and loose drives in one collection saves the call-out and the admin.

A price indication with three scenarios

Here are three typical scales. The exact price depends on the number, the level and whether whole units or only carriers come along, so always request a quote for a fixed price.

Scenario	Scale	What plays a part
Single rack	1 rack with servers, a switch and some loose drives	One collection moment, inventory list helps, possibly with paper
Partial floor	Several racks with storage, tapes and network gear	Lower price per item, planning on site, mixed streams
Whole server room	Full decommissioning, dozens to hundreds of drives	Sharpest price per item, logistics in advance, possibly several trips

If you have only a single rack, it is often economical to have it collected together with paper or other data carriers. For a whole server room we plan the logistics in advance.

How the collection service works

The collection service works in a few clear steps.

1. You request a quote with the number and type of devices and the level you want.
2. We schedule the collection at your location, at a time that suits you.
3. The carriers and units go in a sealed bin, under a registered handover.
4. Everything is shredded to the agreed DIN level.
5. You receive a certificate with the serial numbers of all destroyed carriers and units.

The whole chain is closed, from your server room to the shredder. You do not have to dismantle or transport the equipment yourself, and we handle the removal of the drives during the collection if you wish.

The chain of custody from rack to shredder

A decommissioning is all about the chain of custody, the closed chain in which every carrier stays traceable. From the moment a component comes out of the rack until destruction there is a watertight record. The units go in a sealed cart, the handover is logged and at the shredder every carrier is checked off by serial number, so nothing disappears unnoticed. That closed chain is exactly the difference between handing equipment to a random collector and professional data destruction.

Do not forget the network gear

With network equipment destruction people do not always think of data, yet it belongs there. Switches, routers and firewalls hold configurations, VPN keys and certificates in flash memory. A discarded firewall with the full network configuration is a gift to anyone with bad intentions. That is why we take network gear along in the same collection and destroy the memory modules to the right level. Patch panels, KVM switches and management cards can hold sensitive information too, so make no distinction between servers and network gear.

Tapes and loose media in the same run

A tape library often holds years of full backups. Those tapes belong in the same decommissioning, as do loose drives, NAS devices, USB sticks and memory cards from drawers. The advantage of combining is that you pay for the call-out only once and get a single certificate listing all the streams, so no forgotten tape or stick is left behind. Deep dives per type are in backup tape destruction and having a hard drive shredded.

The data destruction certificate

The proof that everything went well is the data destruction certificate. It states the date, the DIN level applied and the serial number of every destroyed carrier and unit. That serial number ties the proof to your specific equipment, so it lines up with the asset register and lets you show in an audit that exactly that drive or switch was destroyed. What belongs on such a certificate is in data destruction certificate explained.

Demonstrable destruction for the GDPR

The GDPR asks not only that you destroy personal data, but also that you can demonstrate it. A decommissioned storage array left standing somewhere is not destruction, even if nobody uses it. Only once the carriers are irreversibly destroyed and you have a certificate do you meet the requirement of demonstrability. A datacenter often holds the central data of the whole organisation, so the risk is greatest there. A conclusive certificate with serial numbers proves that you take your duty of care seriously.

Datacenter hardware destruction in Amsterdam and nationwide

We collect equipment within a 20 km radius of Amsterdam, with no call-out fees. Whether you are in Amsterdam-Noord, Amstelveen, Zaandam, Diemen or Haarlem, we come to you. Short lines mean fast planning, even if your server room has to be empty by a tight date. You do not have to drop anything off yourself, which is unwise with sensitive data, and the equipment stays under sealed handover until it is destroyed. For anyone outside the immediate region, collection is possible nationwide via pooled routes. Give your postcode with the request and you will know straight away whether you fall within the service area with no call-out fees.

Recycling of materials in datacenters

Destruction and sustainability go together. Servers, racks and storage contain a lot of steel, aluminium, copper and precious metals. After grinding, the metal particles are separated and recovered by a metal processor and get a second life as raw material. Recycling of materials in datacenters means the chassis and racks without data simply enter the recycling stream, while the carriers are ground unreadable. So you combine secure data destruction with circular processing. Batteries from UPS systems are disposed of separately.

Common mistakes

• Destroying only the servers. Tapes, firewalls and loose drives are then left out of sight.
• Forgetting network gear. Switches and firewalls hold configurations and keys.
• Making no inventory. Without an asset register you cannot check anything afterwards.
• Destroying SSDs at too coarse a level. Choose H-5, so the chips really become unreadable.
• Returning complete lease equipment with data on it. Pull the carriers and destroy them separately.
• Not keeping the certificate. Without proof you can demonstrate nothing in an inspection.

Step by step to a safe decommissioning

1. Inventory all the equipment and note the serial numbers in an asset register.
2. Decide the approach, pull-and-shred or whole-unit per type.
3. Decide the level, H-4 for HDD or H-5 for SSD and special data.
4. Request a quote with the number and type of devices.
5. Schedule the collection at your location within the service area.
6. Keep the certificate with your GDPR records.

An example from practice

Suppose an IT manager closes a server room after a cloud migration. There are two racks with twelve rack servers, a storage array with forty drives on SSD, four switches, a firewall and a tape library with old backups. Some servers no longer boot, so wiping is not an option. The manager makes an asset register and requests a quote for level H-5, because the array runs on SSD. Within a working day there is a fixed price per carrier and per unit, with no call-out fees because the office is in Amsterdam. On the agreed day the drives are removed. All the carriers, the tapes and the firewall go along in a sealed bin, and the empty chassis and racks go to recycling. A few days later the certificate is in the inbox, with each serial number listed. The risk is gone and everything is demonstrably in order.

Frequently asked questions

What is covered by datacenter hardware destruction?

Everything that can hold data or configuration: servers, storage arrays and SANs, switches, routers and firewalls, backup tapes, loose drives and SSDs, plus racks and PDUs as a waste stream. Each component is registered by serial number and destroyed to the agreed DIN level.

Do the drives come out or can the whole unit go through the shredder?

Both are possible. With pull-and-shred we remove only the carriers and the chassis goes back or to recycling. With whole-unit the complete unit goes through the shredder, so no forgotten cache module is left behind.

Do you destroy network gear and tapes too?

Yes. Switches, routers, firewalls and backup tapes go along in the same closed chain. We destroy the memory modules and the tapes to the right level, so no configuration or backup is left behind.

What determines the cost of datacenter disposal?

The number of carriers and units, whether you destroy whole units or only the carriers, the level you want and the distance. You pay per item by serial number with a fixed price in advance and no call-out fees within 20 km of Amsterdam.

Do I get a conclusive certificate?

Yes. You receive a data destruction certificate with the date, the DIN level applied and the serial number of every destroyed carrier and unit, so the inventory list reconciles.

Can a whole server room be done in one go?

Yes. For a complete decommissioning we plan the logistics in advance. The removal, the haul-away and the destruction run in a closed chain, with a single certificate for all carriers and possibly several trips.

Conclusion

Datacenter hardware destruction goes beyond the servers alone. The risk is spread across storage, switches, firewalls, tapes and loose drives, so every component belongs in the same closed chain. With a clean inventory, physical shredding at the right DIN level and billing per item by serial number you know for sure no sector or chip is readable any more. The certificate makes it demonstrable afterwards that exactly your equipment was destroyed, so a decommissioning combines maximum security with conclusive proof and responsible recycling.

---

A price straight away? Request a quote via desnipperaar.nl or first read the deep dive on server destruction at end-of-life. You get a fixed price per item in advance, with no obligation.