Home / Data recovery / Linux data recovery
Linux serversDeleted files, a filesystem that will not mount, a RAID that will not assemble, or a volume group that disappeared. We copy the disk first and work on the copy. Your original disk is not touched.
Software problems on disks that can still be read.
Someone ran rm on the wrong folder. We check backups and snapshots first, then scan the copy of the disk for what is left.
After a crash or power cut ext4 or XFS shows errors, drops to read-only, or fails at boot. We repair it on the copy, not on the original.
A disk dropped out, the array shows inactive, or it will not assemble. We read the RAID metadata and rebuild it from copies of the members.
The volume group or a logical volume is gone after a resize or a wrong command. We restore the LVM metadata and bring the volume back.
The partition table was overwritten or the wrong disk was formatted. We look for the old layout and, where it is still there, restore it.
A dropped MySQL or MariaDB database, a lost data folder, or a deleted cPanel account. We check the backups and, where possible, rebuild from the files.
An honest view of each situation. It is not a promise: the result depends on your disk.
Tap a box to show only that group.
Sometimes part of the files can be brought back from the journal or by scanning the free space. Names and folders may be lost.
New data has probably overwritten the old. A backup or snapshot is the better hope.
XFS has no reliable undelete. A scan can find some files by type, but a backup or snapshot is usually the only real way.
This is the best case. Restoring from the snapshot or backup is fast and safe.
Most ext4 and XFS errors can be repaired or read from a copy of the disk.
A careful copy often saves most of the data. The files that sat on the bad areas may be damaged or lost.
If the partition was not formatted afterwards, the old layout can usually be found and put back.
Some files can often be found by scanning the disk, but folder names and structure are usually lost.
The data is usually intact on the other members. We copy them first and assemble from the copies.
It depends on which disks failed and when. Partial recovery is possible, a full recovery is not certain.
LVM keeps old copies of its settings. If the disks were not overwritten, the volume can usually be restored.
Often most of the data can be dumped out. Backups and binary logs give much better results.
Good if a panel or server backup exists. Without a backup, only part of the files may be found.
Without the key, encrypted files cannot be opened. Backups, snapshots and free decryptors for some known families are the only ways. We do not break encryption.
Only possible if the provider still has a snapshot or backup. Check with them at once.
The drive clears deleted data soon after deletion, so recovery chances drop fast.
Overwritten data cannot be recovered by anyone.
This is hardware damage. It needs a clean-room lab, and we do not do that work. Switch the disk off.
The same six things on every job.
The more the disk is used after the loss, the less is left. Stop at once.
HDDs keep deleted data for a long time. SSDs with TRIM, and most cloud disks, often clear it fast.
ext4 sometimes gives files back, XFS rarely. Btrfs, ZFS and LVM snapshots can make it easy.
A healthy disk is easy. A failing disk must be copied carefully, and parts may be lost.
These beat every tool. Check them first.
Repairs, reinstalls, formats and new data on the original disk can end the chance.
The same careful order every time.
Stop the apps that write to the disk. Tell us the distribution, the filesystem, what was lost and when.
We make an image with ddrescue on storage you give us or we arrange, and keep the original as it is.
Snapshots, hosting panel backups, database dumps and RootManage backups are the fastest and safest way back.
We mount the image read-only, repair what can be repaired, and scan for deleted files.
You get a list of what was found. You check it and choose what to take back.
Files go back over a secure link or to a place you choose. We delete our working copies when you confirm.
Well-known tools, used on a copy of your disk.
Copies a failing disk to an image, skips bad areas first and comes back to them. The first step on every job.
TestDisk repairs partition tables and finds lost partitions. PhotoRec carves files out of free space by type, but the original file names are lost.
Try to bring back deleted files on ext3 and ext4 using the journal. The results depend on how much time has passed.
Repair XFS and ext filesystems. We run them on the copy, because a repair can change data.
Examines and reassembles software RAID. We work on copies of the member disks.
vgcfgrestore, pvs, vgs and lvs to read and restore volume group metadata.
Sometimes, but not always. On ext4 and XFS deleted files are often gone for good. We check backups and snapshots first, then scan a copy of the disk. We tell you honestly what looks possible.
No. We copy the disk with ddrescue and work on the copy, so a mistake cannot make things worse.
Often yes, if the RAID metadata and enough member disks are readable. We copy the members first and rebuild on the copies.
Often yes. LVM keeps old metadata copies, so we can usually restore a volume group that was removed by mistake, as long as the disk was not overwritten.
Check the provider snapshots first. If there is none, we can work from an image you can attach to a second server, or through temporary access you give us.
We look for what is still there and what can be restored from backups. Removing the malware is a separate job, which we also do.
It depends on the size of the disk and the problem. Send us the details and we tell you before any paid work starts.
Send the system, what was lost and when. We reply with the next step.