RAID 5 Recovery Walkthrough — Delhi Data Recovery Lab

A RAID array is supposed to survive a disk failure. When it does not, the situation is
usually worse than a single dead drive, because the data is spread across every disk in the
set and nothing is readable on its own.

This is a walkthrough of what a RAID 5 recovery actually involves at CNC’s lab in West
Patel Nagar — what we do, in what order, what it costs, and the one mistake that turns a
recoverable array into a lost one. The scenario below is representative of the RAID
jobs we handle; it is not one named client.

The situation

Table of Contents

A Delhi office runs a four-disk RAID 5 on a small rack server: accounts data, Tally
backups, and eight years of project files. One disk fails. The array keeps running in
degraded mode, as it is designed to. Nobody notices, because nothing looks broken.

Three weeks later a second disk fails. RAID 5 survives one failure, not two. The volume
goes offline and the server will not boot.

The mistake that costs the data

At this point the common reaction is to put in a new disk and tell the controller to
rebuild. Do not do this.

A rebuild writes across every disk in the array. If the remaining disks have bad sectors —
and after a failure they usually do — the rebuild fails partway and overwrites parity data
that a recovery lab needed. We have seen fully recoverable arrays reduced to partial
recoveries by one well-intentioned rebuild attempt.

Power the server down and leave it. That single decision is worth more than anything else
in this article.

Step 1 — Image every disk before touching anything

Nothing is analysed on the original disks. Each one is imaged sector by sector to a
separate healthy drive, using hardware that can read past bad sectors without hammering a
failing disk. Disks with head or motor damage go to the bench first for physical work.

This step alone can take two to five days on a four-disk array. It is also the step that
makes everything afterwards safe — from here on, every attempt runs against a copy.

Step 2 — Rebuild the array on paper

A RAID set has parameters: disk order, block size, parity rotation, start offset. The
controller knew them. With the controller dead or the configuration lost, they have to be
worked out from the data itself by testing combinations until file structures line up.

This is the part that separates a lab from a repair counter. It is analysis, not software
— no consumer tool guesses these reliably.

Step 3 — Mount, verify, hand over

Once the parameters are right, the file system appears and files can be pulled. You see a
file listing before paying anything. Recovered data goes back on a fresh drive, never on the
original disks.

What RAID recovery costs in Delhi

Array type	Indicative cost	Typical turnaround
RAID 0, 2-drive failure	₹8,000 – ₹18,000	10–20 working days
RAID 5, 3–8 drives, single failure	₹10,000 – ₹25,000	10–20 working days
RAID 6, dual-drive failure	₹15,000 – ₹35,000	10–20 working days
RAID 10	₹12,000 – ₹28,000	10–20 working days
NAS — Synology, QNAP, WD My Cloud	₹10,000 – ₹30,000	8–16 working days
Dell / HP / IBM server RAID	₹20,000 – ₹80,000+	Quoted after diagnosis

Diagnosis is free, the quote is confirmed before any work starts, and CNC’s
No Data – No Charge policy applies: if the data does not come back,
you pay nothing. Full details on the
data recovery page.

What would have prevented it

Monitoring. The first disk failed three weeks before anyone knew. A
₹0 email alert from the controller would have turned this into a routine disk swap.
A backup that leaves the array. RAID is redundancy, not backup. It does
not protect against two failures, ransomware, or deletion.
Replacing disks as a set. Drives bought together tend to fail together.

If you are running a server or NAS without monitoring, an
IT AMC covers exactly this — health checks and backup
verification on a schedule rather than after the event.

Frequently asked questions

Can data be recovered after a failed RAID rebuild?

Often yes, but less of it. A failed rebuild overwrites some of what a lab would have used.
The honest answer after a rebuild attempt is that recovery becomes partial more often.

How long does RAID recovery take?

Typically 10–20 working days for a standard array. Most of that is imaging, which
cannot be rushed without risking the disks. Emergency fast-track is available for
business-critical data.

Do you need all the disks from the array?

Bring every disk, including the one that failed first and any spares. Parameters are often
recovered from the disk everybody assumes is useless.

Is RAID recovery possible if the controller card is dead?

Yes. The array is reconstructed from the disks themselves, so a dead controller does not
prevent recovery. Bring the card if you have it — it can shorten the analysis.

What does No Data – No Charge mean here?

If CNC cannot recover your data, you pay nothing — not even the diagnosis. It applies to
RAID and NAS jobs the same as single drives.

Prices and rates mentioned in this article are indicative estimates as of October 2026, exclusive of GST unless stated, and may change without notice due to exchange-rate movements, supply constraints and market demand. Please confirm current pricing and availability with CNC before making any decision.