RAID Array in Degraded Mode: Urgent Steps to Save Data
RAID Array in Degraded Mode: Urgent Steps to Save Data
Is your RAID controller showing a "DEGRADED" status? Is the management interface glowing orange or red? You may have hours – perhaps only minutes – before the situation gets dramatically worse.
A degraded state means that one drive has failed and the array is running without its full redundancy. The array still works, but another failure means data loss.
What Degraded Status Means
Definition
A RAID array is degraded when one or more drives have failed but the number of failures has not yet exceeded what the RAID configuration can tolerate:
| RAID type | Tolerates | Degraded after |
|---|---|---|
| RAID 1 | 1 drive | 1 failure |
| RAID 5 | 1 drive | 1 failure |
| RAID 6 | 2 drives | 1–2 failures |
| RAID 10 | 1 per mirror | 1 failure in a pair |
How the Array Keeps Working
When data from the failed drive is requested, the array calculates the missing data from parity (RAID 5/6) or reads it from the mirror (RAID 1/10). This works, but:
- It is slower
- It puts extra stress on the remaining drives
- Another failure = catastrophe
Why It's a Critical State
No reserve: In a degraded RAID 5, a single bad sector on one of the remaining drives means data loss.
Increased load: The remaining drives compensate for the failed one. More work = a higher risk of another failure.
Domino effect: Drives from the same batch are the same age. If one has failed, the others are probably not far behind.
How Quickly to Act
Risk Timeline
First hours: The array works, but every minute of operation increases the risk. The remaining drives are under stress.
Days: The risk of another failure keeps growing. Drives of the same model and age tend to have a similar lifespan – if one has gone, another may soon follow.
Weeks/months: The company ignores the warnings. "It still works, after all." Until the moment it doesn't.
Rule of Thumb
The older the drives, the faster you must act. An array with new drives gives you more time. An array with five-year-old drives is a ticking time bomb.
NEVER Do These Things
1. Don't Replace Several Drives at Once
Why people do it: "One drive has already failed, so I'll replace all the old ones while I'm at it."
What happens:
- You remove several drives
- The controller loses information
- The array may be initialised (= erased)
- All data is lost
Correct: Replace only the one failed drive. Wait for the rebuild to finish. Only then consider replacing another.
2. Don't Force a Rebuild
What "Force Rebuild" is: A command that makes the controller start a rebuild despite warnings.
When it destroys data:
- When the controller doesn't know which drive holds the current data
- When the metadata is corrupted
- When the failed drive has been misidentified
Correct: Unless you are certain what you are doing, don't force a rebuild. Contact an expert.
3. Don't Initialise the Array
Initialise vs rebuild:
- Rebuild: restores data from parity onto a new drive
- Initialise: creates an empty array and erases everything
Why it happens: The buttons sit close together in the interface. One click decides the fate of your data.
Correct: Check three times before every click. If in doubt, don't click.
4. Don't Disconnect Any Other Drives
Why people do it: "I'll pull the drive out and put it back in – maybe that will help."
What happens:
- The controller loses sync
- Drives may get mixed up
- Metadata may be corrupted
Correct: Leave the drives where they are. Document the state. Get help.
5. Don't Install Recovery Software on the Array
Why it doesn't work: Recovery software is designed for individual drives, not RAID arrays. It cannot interpret striping and parity.
What it can make worse: The software may cause additional writes to the array, which can overwrite data.
Correct: Run recovery software only on sector copies of the drives, never on the live array.
What TO DO
Step 1: Stop Operations
- Inform users about the outage
- Shut down the applications that use the RAID
- Minimise I/O on the array
- Don't power off the server yet (data in the controller's RAM would be lost)
Step 2: Document
Photograph:
- The drive status LEDs
- The management interface
- The event logs
Write down:
- What happened before the failure
- The exact time
- Any error messages
This is critical for diagnostics and any subsequent recovery.
Step 3: Back Up What You Can
If the array is still readable:
- Prioritise the most important data
- Copy it to external storage
- Don't copy everything at once (too much load)
Caution: Copying stresses the remaining drives. Weigh the risk of another failure against the value of the backup.
Step 4: Contact an Expert
What to tell us:
- RAID type (0, 1, 5, 6, 10)
- Number and capacity of the drives
- Controller model
- What happened and when
- How critical the data is
What to prepare:
- Access to the server (physical or remote)
- A contact person in IT
- Who has the authority to approve costs
Can I Keep Running a Degraded RAID?
Short term (hours): possible
If you must finish a critical process, a degraded RAID can keep running. But:
- Minimise the load
- Monitor the state
- Be prepared for failure
Long term: NO
Risks of carrying on:
Another drive failure: one bad sector on the remaining drives = data loss
Overheating: the remaining drives work harder and generate more heat
Power failure: in a degraded state the array is more vulnerable
Psychological trap: "It still works, after all" – until the moment it doesn't
Monitoring and Prevention
SMART Monitoring
Monitor the SMART values of all drives:
- Reallocated Sector Count: growing = the drive is dying
- Current Pending Sector: non-zero = problem
- Spin Retry Count: non-zero = mechanical problem
Alerting
Set up notifications for:
- Degraded status
- SMART warnings
- High drive temperature
- Unusual entries in the event logs
Hot Spare
A drive connected to the array but not in use. When a drive fails, the hot spare automatically takes its place and the rebuild starts.
Advantages:
- Automatic response
- Less time in a degraded state
Disadvantages:
- The rebuild is still risky
- The cost of an idle drive
Regular Checks
- Monthly RAID status check
- Quarterly SMART value check
- Annual review of configuration and capacity
Case Study
Situation
A medium-sized company with an 8-drive RAID 5 in its file server. In use for four years. One drive failed.
What Happened
The IT technician saw "RAID Degraded" and ordered a new drive. But since the server "still worked", nobody hurried. The drive was due to arrive in five days.
Day 4: a second drive failed. The data was lost.
What They Should Have Done
- Minimise operations immediately
- Back up critical data to an external drive
- Order the replacement drive with express delivery
- Consider professional help for a safe rebuild
Lessons
- Degraded status = an urgent state
- Time works against you
- Four-year-old drives are in the risk zone
- Express delivery costs a fraction of what lost data costs
FAQ
How long can a RAID run in a degraded state?
Technically, indefinitely. In practice, the longer it runs, the higher the risk. We recommend resolving it within hours, not days.
Can I replace the drive myself?
If you have experience and are confident: yes. The key points are:
- Identify the failed drive correctly
- Use a compatible replacement drive
- Don't choose "Initialise" instead of "Rebuild"
If you are unsure, contact us first.
What if another drive fails?
RAID 5: data loss (no redundancy left) RAID 6: still works, but in a very critical state RAID 10: depends on which drive (a different mirror pair = OK)
Need Help?
If your RAID is in a degraded state and you are unsure about the next steps, describe your case in writing – the initial assessment is free and you receive a binding quote before any recovery work starts. No Data, No Fee. For business-critical systems we offer Express processing (12–48 h).