NAS principles

The modern NAS device from any vendor is just a small server computer with one or more disk storages attached. Even small NAS device contains specialized board with processer that runs embedded OS. The OS provides support of network file sharing, NAS configuration, file system manipulations and so on.

Unlike single-disk NAS, multi-disk models combine internal storages into RAID to extend disk storage space, increase file access speed or add storage redundancy. The RAID supported in most NAS (Iomega, Synology, Buffalo) is just a software function of embedded OS. Other NAS boxes (like models from Promise) can also use hardware RAID.

The most popular embedded operating systems for NAS are compact Linux OS editions. In this case native NAS file system is Ext3 or XFS. Other vendors (like Adaptec) can install BSD-based solutions (like SnapOS) and use custom UFS editions for main storage. The modern NAS can also use modern ZFS file system editions. The NAS native file system does not affect network files access (each NAS just publishes its file system as 'network file system') in any way, however, affects data recovery.


Prepare for recovery

Irrespective of NAS data loss type: system failure, storage format, file deletion etc. the main step is to connect NAS drives to a computer. Even if NAS works and can boot, it does not provide software with low-level interfaces that makes possible direct drive access, required for data recovery purposes. The only solution is to eject drives from NAS and connect them directly to a computer.

Depending on NAS model and drives number, it may require different kinds of adaptors on PC. Please, refer to our knowledge base article "HOW TO: CONNECT IDE/SATA DRIVES TO RECOVERY PC" for more information about this step. The key point here is to remember ejected drives order for multi-disk NAS: this will help with RAID virtual reconstruction and will help to assemble back the NAS properly. It's recommended to mark drive numbers with paper stickers or with soft ink marker.


Data Recovery

The first step of data recovery itself is to assemble the data storage. Single disk NAS usually contains 'data file system' on large data partition of the disk. The same applies to multi-disk NAS in pure RAID1 configuration.

The multi-disk NAS of Buffalo, Iomega, Synology etc. have software RAID, built on the largest, 'data' partitions of each disk. The NAS systems like Promise NAS use hardware RAID, built on entire disks. Please refer to specific articles about RAID virtual reconstruction for these systems (e.g. HOW TO: DATA RECOVERY FROM BUFFALO TERASTATION).

After data storage is built, the next step is to identify NAS file system and do the required data recovery operations. The Promise or Intel NAS systems can have non-standard extensions to disk partitions so file system may not be detected automatically. That's why short "Scan for lost file systems" may be required.

Having file system opened in UFS Explorer, you may get you data from the NAS: you may just copy data out (in case of hardware failure), reconstruct the file system (in case of file system damage), scan file system for deleted files and so on.