NAS Data Recovery Aberdeen

It was designed for, and it is survivable, but this is also the point where your array is one hiccup away from actual trouble. Consider this light a fire alarm; do not consider it something to check "next week".

It is the resilver load that causes a tired survivor to die. This is by far the most frequent type of NAS disaster that has been booked into this workshop, and not by a small margin.

Array logic has failed. However, the data contained beneath the failing array logic has not failed. This is reconstruction work. Pressing the restore button located on the front panel does nothing towards restoring the original data.

A simple reassuring click resets the configuration back to its default values and also begins overwriting your data on certain models. While good intent exists here, the process should be stopped immediately.

The firmware expired during a flash process around an array which was in pristine condition. The drives have no idea that anything is wrong and don't even care. They come out and are placed in the imaging bays.

Power loss while writing parity means both disk member volumes have different data and are unable to be mounted together. However, both can be reconciled from complete images, which makes the disagreement "something to look at" rather than "something to be afraid of."

Box electronics fail around perfectly good disks much more frequently than the disks themselves fail. The imaged members bring the volume back to life virtually, and the enclosure was only ever the replaceable component.

Both can be carved back for as long as nobody else claims their blocks, so there's only one thing that matters, which is time. Move fast and don't touch anything during this time frame.

Any item accessible via the internet is considered an attractive target by potential attackers. Since attackers typically delete everything they can view, the tree(s) containing snapshots underneath these deleted items may still exist.

Disks pulled, shuffled, second-guessed and reinserted into the array in a hurry, sometimes multiple times. The metadata of this array still contains the correct order of where each disk belongs (even though no one in the room can tell you), as long as the experimenting stops right now.

Since SHR combines multiple sized disks into a single pool, and since both mdadm and LVM sit below it, and btrfs or ext4 sits above it, losing power in the middle of performing a task on this stack causes the entire stack to tear apart. On a DS220+ or a DS923+ we rebuild this stack layer-by-layer from full images, starting with the lowest level first.

When a member fails in the middle of writing (and thus causes the pool to lose sight of how much space it has allocated), flexible volumes collapse. First, each drive is fully imaged; then, once complete, the volume is "stood up" virtually.

Even though the box is bricked, a single-bay unit's Linux-formatted disk remains perfectly functional, so it can be mounted natively (locally) and the entire contents simply pulled off, thereby completing the job.

From time to time, Buffalo box enclosures will lose their array configuration. However, this doesn't affect the individual member disks' configurations, which remain intact. As such, the layout of the system may be recovered from the drives themselves (i.e., read directly off the disks) rather than attempting to coax it out of the enclosure.

If you purchase a batch of ST4000VN006 drives, they will wear down uniformly with respect to each other; therefore, the second failure occurs soon after the first failure. The day one goes bad, image all of them rather than replacing a single disk.

The broken capacity expansion that left the pool stretched halfway with a refusal to mount. Using images, we reconstruct this expansion in order (step-by-step) the same way it was first attempted.

WD40EFPX disks take too long to recover after they have experienced an error and are therefore dropped by the controller. However, put this same disk onto hardware that is willing to wait and it will image just fine.

After years locked inside a sealed cabinet running warm, the box loses drives at roughly the same time. The proximity in loss indicates that there are likely other issues with the remainder of the drives, and that you should image the entire group. It does not indicate replacing one drive.

SMR members can handle their duties every day but fail once resilvers begin. This was demonstrated by the industry publicly several years ago. The disks are imaged using shingle-aware pacing prior to beginning the rebuild process.

The hidden bad sector(s) that were accumulating over time on an archive nobody ever scrubbed are presented all at once in just a short period of time, specifically when there isn't enough redundancy left. Imaging it first defuses that.

The box's small operating system partitions are mirrored across each disk. If those partitions become corrupted, the box will refuse to boot; however, the underlying data volume may be perfect. The data is extracted directly from the images.

Put a cache in "read/write" (RW) mode and it holds data which has never been written out to disk. Lose the RW device, or one half of a cache pair, and the volume will refuse to mount, because a portion of it really is missing. We treat the cache as a complete member of the set, and image it as such.

When a NAS fails, most people's first instinct is to place the disks into a different one. Most vendors design their arrays differently and therefore the new box will offer to format/initialise these disks. Once you accept this option, there is no going back, and the drives need a bench before that happens.

Modern NAS file systems require free space just to track changes, so if your volume is full there is no way to delete files to clear some space. Snapshots occupy this same space, making it worse. It gets untangled from an image, without the risk of working live.

Folder-level encryption is provided by both Synology and QNAP. They use a "key" file which they tell users to save somewhere secure. Wherever "secure" happens to be for those users, many times it's the NAS device itself. Without a "key" or a passphrase, the folder remains locked, and we tell you so at the diagnostic stage.