An unresponsive VM, a bad VMDK (virtual disk), or a corrupted snapshot will lock away a VM that has been reduced to files. Most often the data itself is salvageable: the virtual disk or datastore simply needs to be rebuilt with integrity. We recover all platforms: VMware, Hyper-V, Proxmox and all others for organisations throughout Belfast and Northern Ireland, in-house.
If an issue occurs within a VM do not delete snapshots, do not run consolidation, and do not roll-back. All three options have the potential to lose the delta(s) that contain the current version of your data. Do not take additional snapshots. Leave the datastore alone until we can evaluate it. If possible, halt the host so no writes occur.
Virtual machines fail due to issues with their virtual disks, snapshots, datastores and underlying storage systems. The following items represent some of our most common recovery scenarios.
We recover all hypervisors and virtual disk formats in each data store and underlying storage.
All major virtualisation software platforms such as VMware ESXi & vSphere, Hyper-V, Proxmox VE, Citrix XenServer, XCP-ng, VirtualBox, KVM and Nutanix AHV; plus all other common disk formats (e.g., VMDK, VHDX, VHD, VDI & QCOW2), in addition to the most commonly used datastores for each platform (i.e., VMFS, CSV, NFS & vSAN); and also all types of snapshot/checkpoint configurations (including differencing/thin provisioned disk configurations).
Recovery occurs by restoring the layers starting at the bottom (datastore), working upwards through the (virtual disks, snapshots, etc.), then restore all other layers until fully functional again. Working from a copy of the storage, we rebuild or replace any layer that has failed. Once rebuilt/replaced, we will reassemble your VM and contents of your VM, without ever accessing/interacting with the original datastore during this process.
We have to look at all of these things: the hypervisor, the datastore, disk format and snapshot chain. Then after we know what went wrong with those items, we also need to see how well the components beneath them are doing. After all of that information has been collected, we will provide a written quote for our services.
Once the quote has been accepted by you, we take a read-only image of the datastore or disks (and possibly rebuild a failed RAID, NAS or SAN before taking that image) in order to not damage the live storage of the original virtual machines.
From there we rebuild the failed VMFS, CSV or NFS datastore from the image. Once rebuilt, we will be able to find where the original virtual machines were located and which disks belonged to each one.
We repair the VMDK or VHDX, rebuild a snapped or failed snapshot chain, then merge the delta files into one consistent virtual disk.
Once we have successfully restored the disks containing your guests' files back to working order, we recover the files and can either fix the file system structure for the guest operating system(s), or repair their SQL, Exchange database(s) or other databases as necessary.
We check recovered VMs and their contents after successful completion (that the VM opens successfully), and we can show you what's come back prior to returning them to you.
The VMs and their data will be made available as useable virtual disks or as extracted files, either on a storage medium sized to the recovery, or by secure transmission to a location suitable to you.
Whether it was a VM that had been deleted, a corrupt VMDK file, a corrupted snapshot chain, or a datastore failure due to RAID failure, we perform VM recovery across all platforms, reconstructing the broken layer each time using a copy of the storage.
Give us some information regarding what happened to cause this issue, and an engineer will contact you. Usually within a single business day.
We will get back to you soon. If it is urgent, call 028 9002 0144.
First there is no charge for a free assessment. Second, prior to performing any work, we provide a fixed price written quote. The starting cost for recovering one VM, or virtual disk on healthy storage is £800 + VAT. The minimum price to recover a full datastore or host containing many VMs is £1,200. Both prices are based upon the specific case being assessed. Most often clients only pay for services rendered by us once we have successfully recovered their VM data; although a 50% payment is required when repairing a physical disk(s) for replacement parts and labour costs incurred while working on the client's hardware. The balance is only paid on successful completion of the job.
Examples of successful virtual machine recovery using a variety of platforms and faults, with configurations and outcomes illustrated without revealing any customer information.
A virtual machine had been uninstalled but still existed in the form of the data (blocks) that made it. The datastore was imaged, the VMDK for the VM was identified, and then restored back into an operational state.
The problem was that there was no AVHDX delta available, which broke the chain. We needed to merge the different checkpoints together into one consistent VHDX that could be used by the VM. After doing so, we successfully recovered the most current state of the VM.
An event related to storage had destroyed the header for the VMFS (Virtual Machine File System). In order to recover all of the virtual machines associated with this datastore, we first made a copy of it and then rebuilt it from that copy. All of the virtual machines contained within this datastore were recovered.
When the server crashed, it caused the VHDX file for the VM to become unbootable. To resolve this issue, we needed to repair the virtual disk. Once we had done that, we were able to recover into the guest OS and repair the NTFS (New Technology File System), which gave us access to the data.
During a server migration, something went wrong during the process and it left a QCOW2 disk that was only partially written. In other words, the migration failed halfway through. In order to recover all of the guest's data, we rebuilt it from the remaining parts of the original disk.
The RAID array below the datastore was unavailable. We rebuilt the RAID 5 from images, recovered the VMFS and then restored the virtual machines.
Send or drop off your item for a free diagnostic check with a description of the issue. One of our engineers will review the item, then provide you a written quote prior to doing any repairs.
Prior to sending or dropping off your device for a diagnostic test (free), obtain a quotation from us via telephone or using the provided form. Step one: get the device to our Belfast bench. Carefully wrap the device and include your contact information inside the parcel. Then send it over. There is no charge for this diagnostic service. Once we receive it, you will be given a firm written price to confirm whether you wish to go ahead, before any work is done.
Sending it? The use of a tracked, insured delivery method would be recommended. Alternatively, dropping it off can also be done Monday to Friday, 9am–5:30pm; no prior booking required but do ensure that all items are packed accordingly.
Still waiting to send something? Please complete the form describing the failure/fault in your own words and one of our engineers will get back to you with a quotation based upon your individual circumstances.
We will get back to you soon. If it is urgent, call 028 9002 0144.
Most frequently asked questions regarding recovery of virtual machines.
Normally, yes. As long as you don't overwrite the space where the deleted virtual machine resides, there is a good chance. We create an image, locate the VMDK or VHDX, and recover the machine. The sooner you stop writing to that datastore the better.
Normally not. A VM that will not start normally has either a broken snapshot chain, corrupted virtual disk, or problem with the data store. The data itself is usually still there. We rebuild whatever layer failed to function and recover either the machine or files within it.
Yes, it is one of the most common types of jobs we perform for VMs. When a delta is missing or corrupt, the chain will be broken and could appear to roll-back the VM; however, data is generally available in pieces. We reconstruct the chain and merge deltas together into a single consistent disk. Do not delete snapshots or consolidate first.
Yes. Most often, when a virtual disk has been damaged through a crash or hardware failure (such as a failed storage device), it will be possible to repair it. If there is no way to completely restore the virtual disk to its original state, we can recover directly into the virtual machine and extract the files and databases. The file formats VMDK, VHDX, VHD, VDI and QCOW2 may be recovered.
Pricing for VM recovery begins at £800 + VAT for a single virtual machine or single disk located on good, working storage. Recovery of an entire datastore or host containing multiple VMs would begin at £1,200. Each job is priced individually based on the specifics of the job after completion of our free assessment. In most cases, you will only pay us if we successfully recover your data. If the disks supporting the VMs require physical repair in order to recover them, we take a deposit against the cost of parts required and labour costs. The final payment would only be made once your data is successfully recovered.
Typically, yes: you would only pay for successful recovery of your data. Should the disks below the VM(s) require physical repairs to support their recovery, a deposit would cover those parts & the cost of the bench time associated with the repairs; however the balance would still only be billed upon successful recovery of your data. We clearly explain which applies prior to commencing any work.
All of the major ones including VMware ESXi & vSphere, Hyper-V, Proxmox VE, XenServer/XCP-ng, VirtualBox/KVM and Nutanix AHV using VMDK/VHDX/VHD/VDI/QCOW2 disk types supported across various data store types such as VMFS/CSV/NFS/vSAN.
Yes, this represents a large portion of our business. When the RAID/NAS/SAN beneath a datastore fails we reconstruct that base level storage first. Then we continue to recover the datastore and ultimately the VMs residing on it. At both levels we perform similar construction/recovery efforts.
Yes. As soon as we've reassembled the virtual disk we will access the guest machine, extract its files and repair the guest's file systems, SQL Server, Exchange etc. when necessary, with regard to corrupted information. This does not just involve recovering the virtual machine but also includes the data contained within it.
Typically, restoring a single virtual machine from a healthy store environment takes anywhere from 3 to 7 working days. However, restoring an entire datastore or when the original array has to be rebuilt prior to restoring it may take longer i.e. 7-14 days. Our assessment phase generally takes no longer than a couple of working days, and we can prioritise business-critical cases.
Deliver or post removable disks to our Belfast office, fully insured and clearly labelled to indicate which slot they came from. For RAID, NAS or SAN-based VMs send all disks associated with the RAID set. Our office hours are Monday to Friday, 9am-5:30pm. Please provide your contact details and allow us to book this in and issue you a price based on our free initial assessment.
Free initial assessment, fixed written pricing from £800 + VAT and in most cases, you will only have to pay once we have successfully recovered your data. Every hypervisor type and every virtual disk format are recovered in house by us directly from our base in Belfast. Do not delete snapshots or consolidate: leave the datastore intact and simply send the disks to us.