No parity to compute from
RAID 5 and 6 store parity so a missing member can be calculated back. RAID 0 stores nothing extra at all — there is no redundancy anywhere in the set.
Definitional
Data is striped across every member with no parity and no mirror, so a lost disk takes a proportional slice out of every file on the volume. Be careful with anyone offering to rebuild one from parity — there is none.
Nothing is held in reserve on a stripe, so recovery means reading every member including the failed one and reassembling from images. How much returns is decided by how much of the casualty can be read.
Which is why RAID 0 is fast, and why losing one member affects everything rather than a portion.
A RAID 0 divides data into blocks — the stripe size, typically 64KB or 128KB — and writes them across the members in rotation: block one to disk one, block two to disk two, wrapping round. Reads and writes then happen in parallel across all the disks at once, which is where the speed comes from.
The consequence is that no file lives on a single disk. A 1MB document on a four-disk array with 128KB stripes sits across all four members in eight pieces. Lose one disk and you have lost a quarter of that document — and a quarter of everything else. Not a quarter of your files; a quarter of each of them.
That is why RAID 0 failure feels so total. The volume does not lose some data. It loses coherence.
The only route is making the failed member readable again.
RAID 5 and 6 store parity so a missing member can be calculated back. RAID 0 stores nothing extra at all — there is no redundancy anywhere in the set.
This is the actual job. Donor heads, board repair, firmware work — whatever makes that physical disk readable. Then all members image normally.
Small files that fit within a single stripe on a surviving disk can be carved out intact. Large files rarely survive that way.
Severe platter scoring or a controller that cannot be revived means that share of every file is gone. Nothing changes that.
The same discipline as any array, with less margin for error.
There is no redundancy to absorb a mistake. Every write is a write to the only copy.
Stripe order is central to reassembly. It can be derived, and having it saves real time.
Controller options labelled recover or repair usually mean re-create, which writes new metadata over the description of how your data was arranged.
Especially the failed one — it is the whole job. An array without it can only ever be partial.
Once every member is readable, the reconstruction is well understood.
All disks are imaged read-only, with the failed member repaired first if needed. Stripe size shows up as a repeating structural period across the images; disk order is established by testing arrangements until file system structures resolve coherently rather than as noise; the offset is confirmed the same way.
The array is then assembled virtually from the images, entirely outside the controller, and the file system rebuilt from the assembled volume. RAID work in Manchester is from £500 +VAT after a free 48-hour diagnostic, with a 50% deposit where a member needs physical repair.
Only by recovering that physical disk. RAID 0 has no parity and no mirror, so nothing can compute the missing stripes back — the route is making the failed member readable, then imaging the full set.
Because striping spreads each file across all members. A document on a four-disk array sits in pieces on all four, so losing one member removes a quarter of every file rather than a quarter of your files.
Small files that fit entirely within a stripe on a surviving disk can sometimes be carved out. Anything larger will be missing its share, so the result is fragmentary at best.
There is nothing to rebuild — and options labelled recover or repair usually mean re-create, which writes fresh metadata over the record of how your data was arranged. Power it down instead.
For scratch space, editing caches, or anything reproducible where speed matters and loss is an inconvenience. It is not storage and it is not a backup — the failure probability is higher than a single disk, not lower.
From £500 plus VAT after a free 48-hour diagnostic. Where the failed member needs clean-air or board-level work there is a 50% deposit, with the balance due only on success.