Your server crashed. Here's what to do before you make it worse.
A failing server is one of the very few situations where doing the wrong thing in the first 30 minutes can permanently destroy your data. I'm a USAF veteran with CompTIA Security+. If your server is down, stop rebooting it and call (208) 918-0636.
Symptoms that brought you here
- ✓ Server won't boot past BIOS / POST screen
- ✓ Blue screen on Windows Server, instant reboot loop
- ✓ RAID controller showing "degraded" or "failed"
- ✓ Drive bay light is solid red or blinking amber
- ✓ Hyper-V or VMware host won't start guests
- ✓ "Inaccessible boot device" / "BCD missing"
- ✓ Files renamed with weird extensions, ransom note on desktop
- ✓ Server beeping repeatedly, fans spun up, nothing on screen
The 5 most common causes
These cover the most common small-business server failures.
Disk failure (#1 by a wide margin)
Hard drives have a finite lifespan and they don't fail gracefully. Severity depends entirely on your RAID. A single drive in a healthy RAID 1 or RAID 5 is recoverable. A second drive failing during rebuild on RAID 5, or any failure on RAID 0 or no-RAID, is a data emergency. The drive light is red for a reason. Stop using the server.
Power surge corrupted boot sector
A brownout, lightning hit, or failing UPS can corrupt the master boot record or the BCD store, leaving the server unable to boot even though the drives are fine. Symptoms: "Inaccessible Boot Device", "BOOTMGR is missing", or "0xc000000e". Often recoverable by repairing the boot configuration, without touching your data.
Failed Windows update
A Windows Server cumulative update that doesn't finish cleanly leaves the server in an unbootable state, stuck in the dreaded "Failure configuring Windows updates, reverting changes" loop. Common after unexpected reboots during patch installation. Recovery typically involves a Safe Mode boot, rolling back the update, and stabilizing from there.
Ransomware-encrypted system
If you're seeing files renamed with weird extensions (.locked, .crypt, etc.), a ransom note in every folder, or a wallpaper that's been replaced with a payment demand, you're dealing with a security incident, not just a server failure. The recovery process is completely different. Do not power-cycle the box. Do not connect anything to it. Call me.
System drive full
Sounds simple but it kills servers regularly. The C:\ drive fills to 100% from runaway log files, Windows Update cache, or a misconfigured database log. The server reboots, can't write its swap file, and refuses to come up. Recoverable in minutes once you have a console, but invisible to most owners until the box is dead.
First 30 minutes: 3 steps
Stop rebooting the server
This is one of the most common and costly mistakes. If a drive is failing, every reboot reads more bad sectors and amplifies the damage. If the boot sector is corrupted, repeated power cycles risk corrupting the file system on the way down. Power it off cleanly (or pull the plug if it's frozen), and leave it off until you have a plan.
Photograph the error
Take a clear phone photo of every error message, BIOS warning, RAID controller alert, and drive-bay light state. Write down the exact wording of any blue-screen STOP code. This information shortcuts diagnosis dramatically. I can often identify the problem before I touch the box.
Verify your most recent backup
Before anything else, confirm that you have a recent backup AND that it restores. A backup you've never tested is a backup you don't have. If you use Veeam, Datto, or Backblaze, log into the dashboard from a separate machine and check the last successful job. If it's been failing silently for weeks, that's the critical thing to know before anyone touches the failing server.
The mistakes that cost people their data
- Don't run chkdsk on a failing drive. Chkdsk rewrites the file allocation table. On a drive with bad sectors, it overwrites recoverable data with "fixes" that destroy your only chance of forensic recovery. This is data death. Don't do it.
- Don't restore from backup until you know the root cause. If this is ransomware and you wipe + restore without finding the entry point, the attacker comes back and re-encrypts your shiny new server inside a week. Find the cause first.
- Don't initiate a RAID rebuild blindly. Rebuilding a RAID 5 with a second marginal drive will often kill that second drive mid-rebuild, at which point you have a dead array with all data gone. Image first, rebuild second.
- Don't pull drives out of order. If you have to remove drives from a RAID, label which slot each came from. Reinserting in the wrong order can confuse the controller and trigger a destructive reinitialize.
- Don't pay a ransom on impulse. Even if it feels like the fastest way out, paying often doesn't get your data back, may violate sanctions law, and confirms to the attacker that you're a paying target. There are usually better paths. Talk to a pro first.
What Idaho Proper IT does
Server recovery is a discipline. The goal is to maximize the chance of full data recovery, minimize downtime, and figure out the root cause so this doesn't happen again next quarter.
I'm a USAF veteran, hold CompTIA Security+, and have worked in IT support and administration roles since 2022, including hands-on work with Windows Server, Linux, VMware, Hyper-V, and RAID controllers.
The recovery process
Recovery pricing
$250/hr
emergency rate after hours · 1-hour minimum · scope quote usually same day
No charge if I determine in the first 15 minutes that recovery isn't viable from the available media.