My phone buzzed at 6:40 in the morning with an email from a little DS220+ sitting in a friend's closet. Subject line: "Warning, Storage Pool 1 has degraded." Below it, a second message about a drive with a SMART status of "Warning." She called me in a mild panic, convinced every family photo from the last decade was gone.
It was not. The drive lasted another three weeks, which was plenty of time to copy everything off and swap in a fresh disk.
That gap between "warning appeared" and "drive actually dead" is the whole game. A SMART warning is your Synology telling you the disk is starting to misbehave, not that it has already quit. If you act calmly and in the right order, you almost always get out clean.
What a SMART warning actually is
SMART stands for Self-Monitoring, Analysis, and Reporting Technology. Every modern hard drive and SSD runs it internally, tracking dozens of health counters like reallocated sectors, pending sectors, and read error rates.
Your Synology reads those counters through DSM and compares them against thresholds. When one crosses the line, DSM flips that drive's health status from "Healthy" to "Warning" and fires off the notification you just got. You can read more about how the standard works on the SMART Wikipedia page.
A warning is a yellow light, not a red one. The disk is still readable, but it has flagged something that tends to get worse over time, not better.
Step 1: Read the exact warning in Storage Manager
Open DSM in your browser, go to the main menu, and launch Storage Manager. Click on HDD/SSD in the left panel.
You will see each drive listed with a Health Status column. Find the one marked "Warning" and note which bay it lives in, for example "Drive 2." Getting the bay number right now saves you from pulling the wrong disk later.
Click that drive, then open the Health Info tab. This is where the real story is.
The attributes that matter
You do not need to understand all forty-something SMART attributes. Three do most of the talking on a failing spinning disk:
- Reallocated Sector Count (05): bad sectors the drive has already swapped out for spares. A handful can be stable; a number that climbs week over week is a dying drive.
- Current Pending Sector Count (C5): sectors the drive suspects are bad but has not remapped yet. Anything above zero deserves attention.
- Uncorrectable Sector Count (C6): sectors it could not read at all. This is the one that actually loses data.
If the raw value on any of these is climbing, treat the disk as on its way out regardless of what a single test says.
Step 2: Run an extended SMART test
A warning tells you something is wrong; an extended test tells you how wrong. In Storage Manager, select the flagged drive, click the Health Info or three-dot menu, and choose S.M.A.R.T. Test, then pick Extended Test.
The quick test takes a couple of minutes and honestly is not worth much here. The extended test reads the entire surface of the disk and can take anywhere from two to twelve hours depending on capacity. An 8TB drive will sit around six to eight hours in my experience.
When it finishes, the result reads either "Normal" or "Failed." A "Failed" extended test means replace the disk, no debate. A "Normal" result with a standing SMART warning is the tricky middle ground I will cover below.
Step 3: Decide, replace or watch
Here is the honest decision framework I use. The numbers below refer to the raw values in the Health Info tab.
| What you see | What it means | What to do |
|---|---|---|
| Extended test failed | Confirmed surface damage | Replace now |
| Pending or uncorrectable sectors above 0 and rising | Active data loss risk | Replace within days |
| Reallocated count stable for weeks, test normal | Old damage, currently stable | Watch closely, plan replacement |
| Warning cleared after test, all counts at 0 | Likely a transient glitch | Monitor, no rush |
My default bias leans toward replacement. A drive that has already thrown a warning is statistically far more likely to fail again than a drive that never has. The peace of mind is worth the cost of a disk.
Step 4: Replace the drive the safe way
If you are on SHR or a RAID setup with redundancy (SHR, RAID 1, RAID 5, RAID 6), you can swap the disk without losing the volume. If you are on a single-drive Basic or JBOD setup with no redundancy, copy your data off first, because pulling that disk takes the volume with it.
Buy a replacement of equal or larger capacity. I strongly favor NAS-rated drives here: WD Red Plus or Seagate IronWolf in the 4TB to 8TB range typically run 90 to 160 US dollars and are built for the 24/7 vibration and heat of a NAS enclosure.
- Power posture: most modern Synology units are hot-swappable, but if you are nervous, shut the NAS down cleanly first.
- Pull the tray for the exact bay you noted in Step 1. Double-check the number.
- Mount the new drive in the tray and slide it back in.
- In Storage Manager, the pool shows "Degraded" with a Repair option. Click Repair and select the new drive.
- Let it rebuild. This can take several hours to a full day on large drives. Leave it alone until it says "Healthy."
During a rebuild your data has reduced or zero redundancy, so this is the one moment I really mean it about having that backup ready. If a second drive hiccups mid-rebuild, the backup is your only rope.
When the warning is a false alarm
Not every warning is a dying disk. Loose SATA connections, a flaky drive tray, or a firmware quirk can throw a one-time SMART flag that never returns.
If the extended test passes, all the sector counts read zero, and the warning does not come back after a reboot, you probably caught a transient event. Reseat the drive tray firmly and keep an eye on it for a week or two.
That said, I would never clear the warning and forget about it. Synology's own guidance in the DSM Storage Manager knowledge base is worth a read for the exact repair steps on your model.
The habit that saves you
The single best thing you can do is make sure DSM can actually reach you. Open Control Panel > Notification and set up email or push alerts, then send a test message. A SMART warning nobody sees is useless.
Set a recurring extended SMART test to run monthly in Storage Manager's schedule, and check the Health Info tab whenever an alert lands. Do that, and you will keep catching these warnings while there is still a comfortable three-week runway to react, instead of finding out the hard way at 6:40 in the morning.
Frequently asked questions
Can I keep using a Synology drive that has a SMART warning?
Technically yes, the drive still reads and writes. But a drive that has thrown a warning is far more likely to fail soon, so treat it as living on borrowed time. Keep a fresh backup and plan to replace it, especially if pending or uncorrectable sector counts are above zero.
How long does an extended SMART test take on a Synology?
It depends on drive capacity. A 4TB drive often finishes in three to five hours, while an 8TB or larger disk can take eight to twelve hours. Run it overnight when the NAS is idle so it completes faster and does not slow down your other tasks.
Should I replace the drive if the extended test comes back Normal?
Not always, but check the raw sector counts in the Health Info tab first. If reallocated, pending, and uncorrectable counts are all zero and stable, it may have been a transient glitch worth monitoring. If any of those numbers are rising over time, replace the drive even when the test passes.
Will I lose data when I replace the warned drive?
Not if you have redundancy. On SHR, RAID 1, RAID 5, or RAID 6 you can swap the drive and hit Repair to rebuild without touching your files. On a single-drive Basic or JBOD volume there is no redundancy, so copy your data off before pulling the disk.
What SMART attributes should I actually look at?
Focus on Reallocated Sector Count (05), Current Pending Sector Count (C5), and Uncorrectable Sector Count (C6). A small stable reallocated count can be fine, but any climbing value across these three points to a drive that is degrading and should be replaced.
Why did the SMART warning appear if the drive seems to work fine?
SMART flags problems before they become visible failures, which is the whole point. Sometimes it is a genuine early sign of wear, and occasionally it is a loose connection or tray issue. Reseat the drive, rerun the extended test, and watch whether the warning returns.
