RAID data recovery
We recover data from every RAID level and every kind of server and storage system. Our technicians have completed hundreds of RAID recoveries and work quickly and safely to get your data back and your operation running.
Server down? Here is what to do first
If you lose access to the data on a server, turn the array off. The longer a damaged array runs, the more likely it is to do further harm. Recovering a broken RAID is specialized work, so choose who handles it carefully. You rarely get a second attempt.
The most common cause of permanent loss is a well-meant fix: forcing a drive online, rebuilding onto the wrong disk, or re-creating the array. If you are not certain, stop and call (951) 595-8065.
What is RAID?
RAID stands for redundant array of independent disks. Files are not stored on one disk. They are split into blocks and spread across all the drives, usually with parity information that lets the array survive one or two drive failures. Recovery means working out exactly how the array was laid out, including drive order, block size, parity rotation and offsets, and reassembling it from images of each drive.
Systems we recover
- RAID levels
- RAID 0, 1, 5, 6, 10, 50 and 60, JBOD and spanned volumes, and maker-specific variants.
- Hardware controllers
- Dell PERC, HPE Smart Array and MR, Lenovo ServeRAID and ThinkSystem RAID, Broadcom/LSI MegaRAID, Adaptec and Microchip SmartRAID, Areca, HighPoint and Promise.
- Software-defined storage
- ZFS RAID-Z, Linux mdadm and LVM, Btrfs, Windows Storage Spaces and ReFS, Intel VROC and RST, macOS software RAID.
- Drives
- SATA and SAS hard drives, SAS and SATA SSDs, and NVMe drives in U.2, U.3 and E3.S form on tri-mode controllers.
- Virtualization
- VMware ESXi datastores (VMFS) and vSAN, Microsoft Hyper-V (VHDX), Proxmox and KVM, Citrix XenServer.
- Applications
- SQL Server, Oracle, MySQL and PostgreSQL databases, Exchange stores, file shares and backup repositories.
Brand guides: Dell, HP and HPE, IBM and Lenovo, Supermicro and Apple. For network storage see NAS recovery.
Common RAID failures
- RAID controller failure
- Most servers have a single RAID controller, so its failure is a single point of failure. A replacement controller may not recognize the array, or may offer to initialize it.
- Clicking or failed drives
- Server drives are reliable, and once they fail they are prone to media damage. A second failure during a rebuild is the classic way a RAID 5 is lost.
- Missing or lost configuration
- A power surge, flat cache battery or firmware fault corrupts the configuration held in the controller's memory. The drives appear as foreign or unconfigured.
- Multiple disk failure
- More drives drop out than the RAID level tolerates, often because of a backplane, expander, cable or power supply fault and not the drives themselves.
- Server crash and file system damage
- The array is intact, and the volume, datastore or database on it is corrupt after an unclean shutdown.
- Human error
- The wrong drive pulled, the array re-created, a volume deleted or a server reinstalled over live data.
Models we recover, 2013 to 2026
Examples of the model numbers we see most often, grouped by the years each range was sold. Server models are on the front bezel and the pull-out service tag; RAID controller models are shown in the controller BIOS and on the card label.
| Years | Range | Model numbers |
|---|---|---|
| 2009–2015 | LSI MegaRAID 6Gb/s SAS | MegaRAID 9260-4i, MegaRAID 9260-8i, MegaRAID 9266-8i, MegaRAID 9271-4i, MegaRAID 9271-8i |
| 2011–2019 | Adaptec Series 6, 7 and 8 | Adaptec 6405, Adaptec 6805, Adaptec 7805, Adaptec 71605, Adaptec 8805, Adaptec 81605ZQ |
| 2012–2017 | Dell PowerEdge 12th and 13th generationPERC H710 / H710P, then PERC H730 / H730P | R620, R720, R720xd, T620, R630, R730, R730xd, T630 |
| 2012–2017 | HPE ProLiant Gen8 and Gen9Smart Array P420i, then P440ar | DL360p Gen8, DL380p Gen8, ML350p Gen8, DL360 Gen9, DL380 Gen9, ML350 Gen9 |
| 2012–2018 | IBM System x and Lenovo ThinkServerServeRAID M5110 / M5210; System x passed to Lenovo in 2014 | x3550 M4, x3650 M4, x3550 M5, x3650 M5, RD450, RD650 |
| 2012–2026 | Dell PERC controllers | PERC H710, PERC H730, PERC H740P, PERC H755, PERC H965i |
| 2012–2026 | HPE Smart Array controllers | Smart Array P420, Smart Array P440ar, Smart Array P408i-a, MR416i-a |
| 2012–2026 | Supermicro SuperStorageShips with HBAs in IT mode; RAID is often software or ZFS | SSG-6047R-E1R36L, SSG-6048R-E1CR36L, SSG-6049P-E1CR36L, SSG-640P-E1CR36L, SSG-641E-E1CR36L, SSG-542B-E1CR60 |
| 2013–2026 | Broadcom MegaRAID 12Gb/s SAS and tri-mode | MegaRAID 9361-8i, MegaRAID 9460-8i, MegaRAID 9560-8i, MegaRAID 9660-16i |
| 2017–2026 | Dell PowerEdge 14th to 17th generation | R640, R740, R740xd, R650, R750, R660, R760, R770 |
| 2017–2026 | HPE ProLiant Gen10 to Gen12 | DL360 Gen10, DL380 Gen10, DL380 Gen10 Plus, DL380 Gen11, ML350 Gen11, DL380 Gen12 |
| 2017–2026 | Lenovo ThinkSystemThinkSystem RAID 530-8i / 930-8i / 940-8i | SR630, SR650, SR650 V2, SR650 V3, SR650 V4, ST550 |
| 2017–2026 | Adaptec SmartRAID 3100 | SmartRAID 3154-8i, SmartRAID 3152-8i, SmartRAID 3102-8i, SmartRAID 3154-16i |
Not listed? These are examples, not limits. If your model is missing we almost certainly still recover it. Give us the number when you start a case.
Our RAID recovery process
- Submit the drives. Call us or start a case, then ship or bring every drive from the array, labeled in order.
- Analysis and diagnosis. Each drive is tested and imaged. Failed drives are repaired first. We work out the array layout from the images and give you a firm quote.
- Data verification. We rebuild the array virtually, extract the data and check files, virtual machines and databases. You receive a file list to confirm.
- Payment and return. You pay once you have confirmed the result. Data is returned on new storage or by encrypted transfer.
RMLD works with companies and individuals so that no one has to suffer permanent data loss. For emergencies, our team will work around the clock to complete the job in the shortest possible time.