Server or NAS failed? How to brief a data recovery lab
What to stop doing, what to write down, which keys to find and how to rank your files, so the lab returns what your business needs first.
When the office server or NAS fails and the backup turns out to be missing or out of date, the pressure is to get it back fast. What you do in the first hour, and what you tell the recovery lab, often matters as much as the lab's equipment. A clear brief helps the engineers find the right data, check it, and return the files your business needs first. This guide is for the owner or office manager who has to make those calls.
First, stop making it worse
With RAID and NAS failures, a lot of the lasting damage can happen after the first fault, while people try to fix it. Until you have decided on a plan:
- Don't keep restarting it. Every power cycle stresses drives that may already be failing.
- Don't start a rebuild, initialize, or "repair" a volume because the screen suggests it. On a damaged array, those actions can overwrite the information needed to put the data back together.
- Don't swap or reorder drives, and don't put a drive from the array into another computer to "have a look". Windows may offer to format it.
- Don't run disk repair tools such as CHKDSK or fsck on the volume.
- Power it down if any drive is clicking, grinding or beeping, and leave it off.
If staff can't work, move them onto whatever copies exist, such as cloud storage or a recent backup, rather than writing anything new to the failed system.
Write down what happened
Before anyone touches the hardware, capture the facts while they are fresh. A phone photo is fine.
- The device: make and model of the server or NAS, how many drives, and their sizes.
- How it was set up: RAID level if you know it (RAID 1, 5, 6, 10, or the NAS maker's own type), and whether there were several volumes.
- What you saw: screenshots or photos of error messages, warning lights and email alerts, with dates. Alerts from weeks earlier are often the most useful clue.
- The sequence: a power cut, an update, a drive replaced, a rebuild that stopped halfway. Note who did what and when, including anything an IT person or a vendor already tried.
- Drive positions: if drives must be removed, label each one with its bay number before it comes out, and keep them in that order. Never mix them up.
Find the keys and passwords
Encryption is the step that most often turns a recoverable case into an impossible one. If the volume, the server's disks or individual folders were encrypted, the lab will need the key or passphrase. Without it, the data may stay unreadable even when the drives are perfectly readable. Look for:
- BitLocker recovery keys for Windows servers, which may be saved in a Microsoft account, Active Directory or a printed record.
- NAS volume or shared-folder encryption keys and the NAS admin login.
- Passwords for any encrypted backup files or containers stored on the device.
If you don't know whether encryption was used, say so. That is useful information too.
Decide what matters most
This is the part that saves the most time. A server can hold millions of files, and not all of them are equal. Make a short, ranked list:
- Must have: the files that stop the business if they're lost. Be specific: the accounting company file and where it lived, the client folders for active jobs, a database, a virtual machine, a mailbox archive. Give folder paths and file types if you know them.
- Should have: recent project archives, templates, scanned records.
- Nice to have: old archives, personal folders, installers, media nobody has opened in years.
Also note what you already have elsewhere, such as files synced to the cloud, an older backup, or copies on staff laptops. That tells the lab what they don't need to spend time on, and gives you a way to cross-check that recovered files are the right versions.
Include a few test files: names of documents you know well and roughly when they were last changed. Opening those is a quick way for both sides to confirm the recovery is good.
Practical details for the lab
- One contact person who can answer questions and approve decisions quickly.
- How you want the data back: on a new drive big enough for the "must have" list, or copied to a new server, and in what folder structure.
- Confidentiality needs: client, medical or financial records, and any requirements your contracts or regulators set for handling them.
- Keep the original drives until you have checked the recovered data and have a working backup of it.
After the recovery
Once the files are back, fix what let this become a recovery job. Set up a backup that runs off the device, check that alerts reach someone who reads them, and test a restore. Our article on running a test restore drill covers that last part.
Our server data recovery, RAID data recovery and NAS data recovery pages explain the services and how cases are handled. If your server or NAS has just failed, call (213) 986-4855 before trying anything else.