Storage area networks fail rarely and expensively: multi-shelf arrays, fibre channel fabrics, LUNs mapped to hosts that suddenly see nothing. EDR handles SAN work as structured engagements — scoped at intake, NDA’d as standard, engineered on images — because at this scale, improvisation is the only unaffordable thing.
$ edr connecting…
A SAN’s complexity is indirection: physical disks grouped into pools, pools carved into LUNs, LUNs presented over fibre channel or iSCSI to hosts that treat them as local disks — often with a virtualisation layer on top. When shelves fail, controllers disagree or an administrator’s change orphans a LUN, the data is usually intact but the map to it is broken. Recovery is cartography first: image every relevant member, reconstruct the pool and LUN geometry from on-disk metadata, then rebuild the file systems or VMFS/CSV volumes the hosts expected. Multi-terabyte estates are handled in stages with checkpoints you sign off.
SAN jobs start with a conversation, not a courier: array make and model, shelf and disk counts, what changed before the failure, and what the priority dataset is. From that comes a scoped plan and a fixed quote — larger than a single-drive job, itemised so you can see why. Discretion is default: NDAs signed, media handled by our own engineers only, and updates to a named contact. If your storage vendor or IT partner is involved, we work alongside them happily; the bench’s job is the data, not the politics.
This is the standard SAN-recovery brief: vendor support rebuilds the array's health, not its history. Recovery works from the disks' imaged state to reconstruct the LUNs as they were, targeting the delta your backup misses. Bring both — a partial backup plus recovery beats either alone.
Yes — staged: members imaged in prioritised batches, geometry reconstructed, and the critical LUNs recovered first so your urgent data returns while the long tail processes. Scale changes the timeline and the quote's shape, never the discipline.
Routinely — a large share of SAN work is virtualisation storage. Once the LUN is reconstructed, VMFS is repaired and individual VMDKs extracted and verified; the virtual machine page covers what happens inside them.
A free diagnostic first, always — then a fixed written quote before any work begins. No surprises on the invoice, because there's no invoice until you've approved the number.
SAN-specific write-ups are still rare in the case-files nest, but the discipline is identical to the one that saved this RAID 0 server — image every member first, reconstruct nothing live.
Getting a device onto this bench only takes two moves, and the clock on your free look starts the moment it lands.
Strip out cables, caddies and power bricks — none of it helps — wrap the device so it can't shift in transit, and either post it insured to 83 Princes Street, Edinburgh EH2 2ER or bring it in yourself. Tuck a note of what happened in with it, and enclose the shipping form if you've printed one.
Arrival gets confirmed, the diagnostic runs at no cost to you, and the call that follows names the fault and states a fixed price. The bench stays idle on your device until you say go.
Free diagnostic, a written per-array quote, and disks imaged read-only before any reconstruction begins.