How to Replace a Failed Drive in Synology DSM

The beep is what gets people. A Synology unit with a dead drive emits one long continuous tone, and at 2am that sound will convince you the whole box is bricked. It almost never is.

A single failed drive in a redundant volume means your data is still intact and readable. The array is just running without its safety net until you swap in a healthy disk. That is the whole job here: pull one drive, push in another, and click the repair button.

I have done this dozens of times on everything from a two-bay DS220+ sitting on a bookshelf to a stuffed DS1821+ in a home rack. The steps below are the exact order I follow, and I will flag the two or three places where people wreck a working array by moving too fast.

First, confirm the drive actually failed

Before you touch any hardware, log into DSM in a browser and open Storage Manager. Do not trust the front-panel light alone, because a loose SATA connection or a firmware hiccup can flash the same amber LED as a genuinely dead disk.

Inside Storage Manager, look at the HDD/SSD section. A truly failed drive shows a status of Crashed or Failed, and the volume above it will read Degraded. If instead you see Critical or a warning with rising bad sector counts, the drive is dying but has not dropped out yet, which changes your approach.

Write down the bay number DSM reports as failed. Synology counts bays starting from Disk 1, usually left to right or top to bottom depending on the model. On a DS920+ the bays are numbered on the drive trays themselves, so match the number in Storage Manager to the physical label. Pulling the wrong bay is the single most common way people turn a recoverable situation into a real disaster.

Silence the beep and check your SMART data

That continuous alarm helps nobody while you work. In DSM go to Control Panel, then Hardware & Power, and click Beep Off. On newer DSM 7 builds the mute button also appears right in the notification. The alarm returns for the next event, so muting it now costs you nothing.

While you are in Storage Manager, click the failing drive and open its Health Info. Look at the SMART attributes, specifically Reallocated Sector Count and Current Pending Sector. A drive with hundreds of pending sectors is genuinely done, and this confirmation stops you second-guessing yourself later. Backblaze publishes years of real-world failure data that maps these attributes to actual death rates if you want the deeper reasoning behind trusting them, over at their SMART stats breakdown.

Buy the right replacement drive

The replacement must be equal to or larger than the drive it replaces. Smaller will be rejected by the repair process, full stop. If you are running classic RAID (not SHR), a larger drive only counts at the size of the smallest member, so a 12TB replacing a failed 8TB in a RAID 5 array gives you 8TB of usable space until every disk is upsized.

Match the drive class too. A NAS-rated drive like a WD Red Plus or Seagate IronWolf handles the constant vibration and 24/7 duty cycle that a cheap desktop drive does not. I have watched shucked desktop drives fail again within months in a busy four-bay unit.

Pull the failed drive

Most Synology desktop models are hot-swappable, meaning you can replace a drive while the unit is powered on and serving files. The DS220+, DS420+, DS920+, DS1821+ and similar all support this. Only a few older or J-series boxes require a shutdown, and DSM will tell you if yours does.

Here is the physical sequence I use every time.

1. Open the front cover or door on the NAS.
2. Find the tray matching the failed bay number you wrote down. Double-check it against the amber light on that specific tray.
3. Press the release latch or push-clip on the tray, then slide the tray straight out slowly. Give the drive two or three seconds to spin down as it clears the backplane.
4. Set the old drive aside on a non-conductive surface. Do not toss it in a drawer with loose screws.

If the tray has no light and you are unsure, cross-reference one more time in Storage Manager. It is worth the extra minute.

Fit the new drive into the tray

Slide the old drive out of the plastic tray. On tool-less trays you pop off the side rails; on older metal trays you back out four small screws from the underside.

Seat the new drive with the connector end facing the back of the tray, so it mates with the backplane when inserted. Reattach the rails or screws snugly, not gorilla-tight. A loose drive vibrates and throws SMART errors that look like a second failure but are really just a rattle.

Slide the loaded tray back into the empty bay until the latch clicks home. You should feel a small resistance as the connector seats. Firm and steady, never forced.

Start the repair in DSM

Return to your browser and Storage Manager. Within a minute or two DSM detects the new disk and the volume or storage pool shows an option to repair.

Select the degraded Storage Pool, click Repair (older DSM 6 calls it Manage then Repair), and follow the wizard. DSM asks you to pick the newly inserted drive as the repair target. Confirm it is the correct bay, accept that the drive will be fully erased, and start the process.

How long the rebuild takes

This is the part nobody warns you about. Rebuilds are slow, and the bigger the drives, the slower they get. Below are realistic figures I have seen on home units under light use.

Heavy read/write activity during the rebuild stretches these numbers considerably. If you can leave the NAS mostly idle overnight, do it. The array is at its most vulnerable during this window, because a second drive failing before the rebuild finishes means data loss on single-redundancy setups like SHR-1 and RAID 5.

You can watch progress as a percentage in Storage Manager. Resist the urge to reboot or pull anything until it reads Healthy.

After the rebuild finishes

Once the pool status returns to Healthy, run a quick sanity pass. Open the failed drive's replacement in Storage Manager and confirm its SMART status is clean, then browse a few shared folders to make sure files open normally.

If you replaced a smaller drive with a larger one on an SHR volume and every other member is now equal or larger, DSM may offer to expand the volume automatically. That extra space appears without any reformatting, which is one of the genuine perks of Synology Hybrid RAID. Synology documents the exact expansion behavior in their knowledge base article on replacing drives if you want to confirm before committing.

A word on what redundancy is not

Replacing a failed drive restores your safety net. It does not, and never did, protect you from ransomware, a fire, a lightning surge, or you deleting the wrong folder. RAID is uptime, not backup.

Related reading: NAS Hard Drive Failure: Signs, Causes and Fixes, Bad Sectors on a NAS Drive: Should You Worry? and NAS Not Recognizing a New Hard Drive: Fixes.

The drive you just pulled proves hardware fails on its own schedule, usually the worst one. Keep a real backup running on a separate device or off-site, and treat this whole exercise as a reminder rather than a scare. You handled it, the data survived, and your array is whole again.