Computer Disaster Recovery Singapore

Most people searching for computer disaster recovery in Singapore want one of two very different things. Either they want to plan ahead — backups, failover, continuity — or something has already gone wrong and the data is sitting on damaged hardware right now. This page is about the second one.

A flood in a Singapore shophouse basement, a lightning surge through an unprotected power strip, a server room that ran hot over a long weekend, a ransomware payload that walked the network on a Friday night — the hardware afterwards is rarely uniformly dead. Platters usually survive fire that destroys the enclosure. Boards die while the media underneath is intact. Arrays lose their configuration while every member disk still holds its stripe.

That gap between "the computer is destroyed" and "the data is gone" is the whole job. We have run this lab for over 20 years and completed more than 140,000 recoveries, and across all case types our success rate sits between 85% and 95% — though a disaster case genuinely can fall either side of that, and we will tell you which after the diagnosis rather than before it.

Two things decide a disaster case and only one of them is the disaster. The first is what the event did to the media. The second is what happened in the hours afterwards — how many times it was switched on, whether someone dried it, whether a rebuild was started. Nothing can be done about the first. The second is still in your hands while you are reading this, which is why the next section comes before everything else.

More disaster-hit data is lost in the hours after the event than during it. The instinct is to check whether it still works. That instinct is the problem.

• Do not keep powering it on to see if it comes back. Each attempt on a physically damaged drive drags heads across the platters. This is the single most common way a recoverable case becomes an unrecoverable one, and the damage is cumulative: the tenth attempt is not the same as the first.
• Do not dry a wet drive, and do not open it. Water leaves residue as it evaporates and corrosion accelerates in air. A drive that is still damp is usually in better shape than one someone has dried. Keep it sealed and get it to a lab. Singapore's humidity works against you here — a drive left on a bench does not dry, it corrodes slowly for days.
• Do not rebuild or resync a failed array. A rebuild writes across the very disks holding your only copy. A rebuild is not a read operation: it calculates parity and writes it onto a member disk, and if the array's assumptions about disk order or offsets are wrong, it writes that wrong result over real data. If a RAID has dropped members, power it down and label the disks in bay order first.
• Do not run recovery software on the original media. Scanning writes, and it keeps the patient under load. Work from an image, never the original.
• Do not reinstall the operating system to "get it working again". That overwrites exactly the structures a recovery depends on.
• After ransomware, do not wipe and restore until the data is secured. Encrypted files and shadow copies are sometimes partially recoverable, but not once they have been overwritten.
• Do not clean anything — not soot, not flood residue. Smoke residue is corrosive and mildly conductive, and wiping a board pushes it into connectors and under components rather than removing it. Bag it as it is.
• Do not use rice, sunlight, a hairdryer or an oven. None of them reach the sealed interior, and all of them accelerate the corrosion that is the real threat.
• Do not fit a board from an identical drive. Modern drives hold calibration data unique to their own head stack on the board itself, so swapping boards without transferring it can make a readable drive unreadable.
• Do not let a general IT shop open it “just to look”. A head flies a few nanometres above the platter. A fingerprint is thicker than that gap, and so is ordinary room dust.

If the machine is on and behaving oddly, the safest single action is to shut it down cleanly and stop using it.

If it is already off, leave it off. Nothing you learn from one more attempt is worth what the attempt can cost.

Handling differs by damage type, and getting it wrong in the first hour routinely costs more than the disaster itself did.

Wet. Seal it in a bag while it is still damp, with a little of the same water in the bag if you can. Corrosion needs air as well as moisture, so keeping the surfaces wet slows it down. Days matter here; weeks do not end well.

Burnt or smoke-affected. Bag it exactly as found, soot included. The residue is a lab problem; the scratches left by someone wiping it are permanent.

Power surge. Leave it off, and resist trying another power supply, cable or enclosure to see whether it was only the adapter. If the surge reached the drive's own board, pushing power through damaged components again can take out the parts that were still fine. This is usually the most recoverable disaster on the list, and one of the easiest to spoil.

Dropped, crushed or struck. Do not shake it to hear whether it still spins, and do not give it one more try.

Ransomware. Pull the machine off the network rather than shutting it down mid-encryption, and do not reformat or restore over anything until what remains has been secured. See our ransomware recovery page.

Whatever the damage: label things before they move. Arrays need their disks labelled in bay order, and a multi-machine site needs every drive marked with the machine it came out of. Reconstructing that afterwards is guesswork, and guesswork spends hours we would rather spend on your data.

In Singapore the pattern is fairly consistent. Water and power dominate, and both are seasonal.

• Water — burst pipes, aircon condensate over a rack, monsoon flooding in ground-floor and basement units. Usually survivable if it reaches us before corrosion sets in.
• Power surge and lightning — takes out the drive's control board while leaving the platters or NAND untouched. Often one of the better outcomes.
• Fire and heat — the enclosure and electronics go first. Sealed media frequently survives what the chassis does not.
• Ransomware and malware — see our ransomware recovery page for what is and is not realistic.
• RAID and NAS failure — a second disk failing mid-rebuild, a controller dying, or an array that reports its volume as crashed.
• Physical shock and theft recovery — dropped machines, and drives recovered from damaged premises.
• Human error — a migration that overwrote the source, a reformatted volume, a deletion nobody noticed until the backup rotated out.

The mechanism matters more than the label, because it decides what is left to work with.

Fire. Heat destroys in a predictable order: plastics, labels, cables and the circuit board long before the sealed interior. Platters are aluminium or glass with a magnetic coating and NAND is ceramic and silicon, so both often read normally after the outside of the drive is unrecognisable. What usually ends a fire case is not the fire — it is the suppression water, or someone cleaning off the soot.

Power surge. A surge normally finds the weakest component on the way in — a fuse, a protection diode, the motor controller — and stops there. Platters are magnetic and NAND cells hold charge; neither is directly touched by a voltage spike on the interface. That is why board-level repair gets a lot of these back intact, and why it carries the best odds on this list.

Impact and crush. A running drive's heads fly a few nanometres above the platter. Any impact hard enough to close that gap puts head onto surface, and from there the danger is the spinning that follows rather than the moment of contact. This is why “it still spins, I will just copy it off quickly” ends so many cases.

Ransomware. It belongs on a disaster page because it behaves like one: it reaches everything at once, backup shares included. Encryption is not reversible without the key, but three things commonly survive a run — shadow copies that were never cleared, backup targets that happened to be offline, and very large files the payload only partly encrypted before moving on.

Disaster cases rarely arrive as a single tidy drive. They arrive as whatever was in the room.

• Desktops and workstations — internal drives pulled from damaged machines.
• Laptops — including boards soaked or burnt beyond use. See laptop data recovery.
• Servers and RAID arrays — reconstructed without the original controller where necessary. See server recovery and RAID recovery.
• NAS units — Synology, QNAP and others. See NAS recovery.
• External drives, SSDs, USB sticks and memory cards.

Work is carried out in our own cleanroom lab at 6 Harper Road. Media does not leave Singapore for the recovery.

A flooded unit or a burnt server room is not one recovery, it is a queue, and the order you put it in decides how much comes back.

Order by how fast each item is deteriorating, not by how valuable it is. Value decides what you are willing to pay to recover; the clock decides what is still there to recover.

• Wet media first, always. Corrosion is the only damage here that is still getting worse while the hardware sits in a box.
• Then anything still powered and misbehaving. A drive failing while in use adds damage every minute it stays up.
• Then single-copy data. If a file exists in two places and only one of them was in the room, that is an inconvenience, not an emergency.
• Include the backup target. The commonest unpleasant discovery after a site disaster is that the NAS holding the backups was in the same room as the disaster. Check it rather than assuming it is the fallback.
• Dry, undamaged, unpowered drives last. They are stable. They wait.

You do not have to send everything at once, and usually should not. Send the priority set first and keep the rest labelled and dry until you see what that diagnosis says.

Document before anything moves. Photograph the room, the rack and each machine where it stands, then serial numbers and bay order before the disks come out, and keep the damaged enclosures rather than discarding them. Loss adjusters ask what condition things were found in, and once hardware has been moved, cleaned and stripped that evidence is gone. Our written diagnosis and fixed quote are itemised documents you are free to pass on — though we are not loss adjusters and we do not handle claims.

1. Triage, same day where possible. Tell us what happened and what the hardware has been through. Water, fire and surge cases are time-sensitive and we will say so on the call.
2. Free diagnosis. We establish what is physically wrong and what is readable. You get a written finding, not a verbal guess.
3. Fixed written quote. Before any recovery work starts. If we do not believe it is recoverable, we say that instead of charging you to find out slowly.
4. Stabilise, then image. Cleaning, board repair or parts as needed — then a sector-level image. All recovery work happens on the image, never the original.
5. Reconstruct and extract. Array geometry, file systems and databases rebuilt from the image.
6. You verify before you pay. We give you a file listing to check. Payment follows your confirmation, not ours.

One detail about the middle of that list is worth knowing: the image is rarely one clean pass. A damaged drive is read in stages, taking the good regions first and returning to the difficult ones afterwards, because a failing drive has a limited number of working hours in it and the order you spend them in decides how much comes out. Every piece of media is logged by serial number on arrival, so a multi-drive case stays traceable throughout.

Business disasters do not wait for office hours. Call +65 6588 0261 and say it is an emergency — priority handling means the case goes to the front of the queue rather than into the normal order.

We sign NDAs for business customers as a matter of routine, and we do not disclose client names or describe the contents of anyone's files. If your case involves personal data, note that recovery is a processing activity under the PDPA and we handle it accordingly.

If the media is encrypted — BitLocker, FileVault, a self-encrypting drive — recovery requires your own password or recovery key. We cannot bypass encryption, and any lab that says otherwise is telling you something untrue.

Be clear about what emergency handling changes. It changes your position in the queue and how fast an engineer gets to the media. It does not shorten the time a flooded eight-disk array takes to clean, image and reconstruct — that work has its own duration and urgency does not compress it. We would rather say so on the first call than have you find out on day three.

Price follows the damage, not the panic. A surge-damaged drive needing a board repair sits at the lower end. A flooded multi-disk array needing cleanroom work on several members sits at the higher end.

What does not change: the diagnosis is free, the quote is fixed and in writing before work begins, and you verify the recovered files before paying. See data recovery cost for the current ranges.

What moves the number is the amount of physical work: how many pieces of media are involved, whether the drive has to be opened in the cleanroom, whether donor parts are needed to read it, and how much reconstruction the file system or array needs afterwards. A surge-damaged drive and a flooded four-disk NAS are different jobs, not different sizes of the same job.

The one cost driver still under your control is what happens before the media reaches us. Repeated power-on attempts, a started rebuild and DIY recovery attempts all add damage, and added damage adds work.

Bring the hardware to 6 Harper Road, Leong Huat Building #05-02, Singapore 369674 — beside Tai Seng MRT — or call first if it is wet, burnt or business-critical, because how you transport it matters.

You will get a written diagnosis and a fixed quote before anything is charged. If the answer is that it cannot be recovered, that answer is free too.

My office flooded. Is a wet drive worth trying?
Usually yes, and sooner is much better. Do not dry it and do not power it on. Keep it sealed and bring it in — corrosion, not the water itself, is what finally kills these.

The drive was in a fire. Is there any point?
Often, yes. Fire destroys enclosures and electronics long before it reaches sealed media. We have recovered from drives whose casings were visibly damaged.

A power surge killed the machine. Is the data gone?
Frequently not. Surges tend to take the control board while leaving the platters or NAND intact, which is one of the more recoverable failure modes.

Our RAID failed during a rebuild. Can it be saved?
Often. Stop the rebuild, power the unit down and label the disks in bay order. We reconstruct arrays without the original controller where needed.

How long does disaster recovery take?
A straightforward case is typically days. Multi-disk and physically damaged cases take longer. We give you a realistic timeframe with the quote, and emergency handling is available.

Do you guarantee recovery?
No, and nobody honestly can. Our success rate across all case types runs 85–95%, but a disaster case depends on what the event did to the media and what was tried afterwards. The free diagnosis is how you find out where yours stands.

Our office flooded and several machines are affected. Do we send everything at once?
Usually no. Send the wet media and the single-copy data first, keep the rest labelled and dry, and decide the second batch once the first diagnosis comes back.

How should I transport a wet or burnt drive?
Sealed, and in the state you found it. Wet drives travel damp in a sealed bag; burnt ones travel with the soot still on them. Do not clean, dry, or pack anything in absorbent material. Call first if it is business-critical.

Everything was backed up to the NAS and the NAS was in the same room. Is anything left?
Frequently, yes. NAS units are multi-disk and the damage is rarely even across all of them, and we rebuild arrays from images of the individual disks without needing the original unit to work. Bring the enclosure with the disks labelled in bay order.

Will you deal with our insurer?
You get a written diagnosis and an itemised fixed quote, which is normally what an insurer asks for, and you are free to pass them on. We are not loss adjusters: we do not handle claims, assess liability or value your loss.

Can you set up backup for us so this does not happen again?
No. We recover data and that is all we sell. The prevention section above is the advice we would give you anyway, but backup design, failover and continuity planning are somebody else's job.

• Server data recovery
• RAID data recovery
• NAS data recovery
• Ransomware recovery
• Water-damaged hard drive recovery
• What data recovery costs

Roughly half the people searching for disaster recovery want a plan, not a lab. If that is you, here is the honest version, free, from the people who see what actually fails.

Industry surveys of Singapore businesses put the gap plainly: about 83% say they are not fully confident of recovering from a serious disruption, and around 62% report unplanned downtime in the past twelve months. Most of those organisations do have backups. What they do not have is a restore anybody has tested.

From what arrives in this lab, the same four failures repeat:

• The backup was in the same room. A NAS on the shelf beside the server protects you against a disk failing. It protects you against nothing else on this page, because water, fire, theft and ransomware all take the whole room.
• The backup ran but nobody ever restored from it. A job reporting success and a set of files you can open are two different claims.
• The backup share was writable from the machine being backed up. That is the exact path ransomware uses. One copy needs to sit somewhere the production machine cannot reach and rewrite.
• Nobody knew whose job it was. Write down who checks it and how often, even if the answer is one person, once a month.

We do not sell any of that, and the section below says so again. Telling you costs us nothing, and the cheapest recovery is the one you did not need.

We do: recover data from computers, servers, arrays and storage media that have been through fire, flood, power surge, ransomware, physical damage or hardware failure. Cleanroom work, board-level repair, array reconstruction, and extraction to media you can use.

We do not: sell disaster-recovery planning, backup software, failover infrastructure or business-continuity consulting. If you have arrived here looking to prevent the next outage rather than to rescue this one, we are not the right supplier and we would rather say so now.

We also cannot bypass encryption. If the media is encrypted, recovery needs the owner's own password or key. Diagnosis is free, and all work is quoted in writing before it begins.