The short answer: On supported Seagate drives, you can read the actual power-on hours from the FARM log using
smartctl -l farm /dev/sdXin smartmontools version 7.4 or later. This provides a second data source alongside standard SMART—and discrepancies between the two exposed drives sold as new after their SMART values had been reset. Perform this check as soon as you receive the drive, before adding it to a NAS or RAID array: array initialization writes its own metadata to the drive and weakens your position if you need to make a warranty claim.
Why people started checking actual power-on hours
Until January 2025, checking the operating hours of a new drive seemed overly meticulous. Then the German publication Heise began reporting cases of Seagate drives that were sold as new despite having already spent years in continuous operation. Within a few weeks, hundreds of reports emerged—from Germany, Switzerland, Japan, Australia, the United Kingdom, and the United States.
Initially, the issue mainly involved data-center Exos drives, but it later expanded to include IronWolf and IronWolf Pro NAS models ranging from roughly 4 to 16 TB. They all shared the same pattern: standard SMART data showed zero or absurdly few operating hours, while the FARM log on the same drive reported between approximately 15,000 and 50,000 hours. Fifty thousand hours is just under six years of continuous operation.
The investigation traced the drives back to hardware retired from Chinese Chia mining farms—although that is the conclusion of investigative reporting, not proven provenance for every individual drive. For buyers, that distinction is largely beside the point. What matters is that the check takes two minutes and that you do not want to discover the problem after adding the drive to an array.
SMART, Device Statistics, and FARM—three counters, not one
The basic principle behind the entire check is simple: do not rely on a single figure; compare three different data sources and look for discrepancies. They are not truly independent—all three counters are maintained by the drive firmware, may draw on shared data, and could potentially be reset together. Their practical value lies in comparison: three values stored and read in different ways should present a consistent picture.
| Record | Command | What it contains | Limitation |
|---|---|---|---|
| SMART attributes | smartctl -A /dev/sdX | Attribute 9 Power_On_Hours, 12 Power_Cycle_Count, and, on Seagate drives, 240 Head_Flying_Hours | These are the values that were reset in the reported cases |
| Device Statistics | smartctl -l devstat /dev/sdX | Standard ATA log: Power-on Hours, Lifetime Power-On Resets, Logical Sectors Written, and Read | Not available on every model, but works across brands |
| FARM | smartctl -l farm /dev/sdX | Extended Seagate record: hours, head flight hours, assembly date, serial number, and recording type | Only supported Seagate drives, smartmontools 7.4+ |
None of these records is a manufacturer-signed certificate, and there is no guarantee that nobody can alter them. The strength of the method lies not in one number but in the comparison—and a discrepancy between sources is a warning sign, not conclusive proof of tampering. A drive that reports three hours in SMART, twenty thousand in Device Statistics, and a 2019 assembly date in FARM is inconsistent enough that it should not be deployed without an explanation from the seller.
What to look for in the FARM log
FARM stands for Field Accessible Reliability Metrics—an extended telemetry record maintained by the drive itself and readable by smartmontools since version 7.4. In the output, the following entries in the Drive Information section deserve particular attention:
Power on Hours— the total time the drive has been powered on. This is the primary figure.Spindle Power on Hours— the time the platters were spinning. This is usually lower because the drive may temporarily stop spinning. Do not confuse it with the previous entry.Head Flight Hours— the time the heads actually spent flying above the platter surface. A genuinely new drive should show single digits; a heavily used one may show thousands.Power Cycle CountandHead Load Events— how many times the drive powered up and how many times the heads moved onto and off the parking ramp. A high figure on a „new" drive has no reasonable explanation.Assembly Date (YYWW)— the year and week of assembly. Compare it with the date on the label and the purchase date. A substantial gap between assembly and sale does not by itself mean the drive was used—it may simply have sat in storage for years—but it is a reason to examine its operating hours and supply-chain provenance.Serial NumberandModel Number— these must match the label, packaging, and invoice. The serial number is reported by the drive firmware, not by a paper label.Drive Recording Type— CMR or SMR. This is unrelated to the drive’s age, but it is worth knowing what you actually bought; the difference is explained in the CMR and SMR overview.
Step-by-step procedure
Before you begin: do not add the drive yet to a storage pool in Synology DSM or QNAP QTS, create a volume on it in TrueNAS, or connect it to a hardware RAID controller. Array initialization writes its own metadata to the drive, which both alters part of its original state and makes it harder to return a unit that you maintain was never „used."
1. Check the tool version.
smartctl --versionYou need version 7.4 or later—this is the release that added FARM log support for Seagate ATA and SCSI drives. An older build will return an error for -l farm, not an empty result. Pay attention to that distinction so you do not mistake missing support for a clean drive.
2. Let the tool find the drives.
sudo smartctl --scan-openThe output will show specific paths, such as /dev/sda -d sat. Do not enter /dev/sdX literally—X is merely a placeholder used to keep the examples concise.
3. Confirm that you are checking the correct unit.
sudo smartctl -i /dev/sdaCompare the model, capacity, and serial number with the label, packaging, and invoice. When receiving a shipment of eight drives, this is why a single test cannot cover the entire delivery.
4. Read all three counters.
sudo smartctl -A /dev/sda
sudo smartctl -l devstat /dev/sda
sudo smartctl -l farm /dev/sdaIf --scan-open also listed a device type for the drive, retain it:
sudo smartctl -d sat -l farm /dev/sda5. Save the output, not just your impression of what was on the screen.
sudo smartctl -x -l farm /dev/sda > seagate-SN12345678-receipt.txtInclude the serial number in the filename. Add photographs of the label, packaging, and SATA connector. If you later need to dispute the issue with the seller, this file will be your main piece of evidence.
On Windows, install smartmontools, open Command Prompt as an administrator, and use the identifier returned by smartctl --scan-open—the /dev/sda mapping works there as well. Seagate also offers SeaTools, which can display basic drive information, including power-on hours. For a disputed unit, however, you should still save the text output from smartctl because it brings all the counters together in one place.
How to interpret the results
Zero hours is not a requirement. Drives undergo testing at the factory and may legitimately show a few hours of operation; there is no single published limit covering every model and distribution channel. The decisive factor is discrepancy, not the absolute number.
| What you see | Practical conclusion |
|---|---|
| SMART, devstat, and FARM broadly agree, with hours in the single digits | Nothing suggests tampering; proceed with health testing |
| SMART is near zero, but FARM or devstat shows thousands of hours | Classic pattern of reset SMART data. Do not deploy the drive; contact the seller |
| FARM shows tens of thousands of hours | A heavily used unit. This is incompatible with a new drive under any circumstances |
| Assembly date is several years earlier than the purchase date | It may not be fraud—the drive could have sat in storage—but ask for an explanation if it was sold as „new" |
Logical Sectors Written corresponds to hundreds of TB | Data has passed through the drive that could not have originated from factory testing |
| The serial number in the output does not match the label | Stop the installation regardless of the number of hours |
| FARM log is unavailable | Inconclusive. It does not confirm that the drive is new; it only means the information is missing |
The last row is where people most often make a mistake. A missing FARM log is not a green light. It is a gap that must be filled with other evidence.
Why FARM sometimes cannot be read
FARM is a manufacturer-specific extension, not a standard ATA command. The entire communication path between the drive and the computer must therefore pass it through correctly—and three situations commonly prevent that:
- A cheap USB-to-SATA bridge. It may not pass vendor-specific ATA commands, causing
smartctlto return an error or meaningless output. Try the device type recommended by--scan-open, or use-d sat,-d sat,12,-d usbjmicron, or-d usbprolific. If none of these works, you need a different bridge with ATA passthrough—or a direct SATA connection. - A hardware RAID controller. It presents a virtual volume instead of individual drives, leaving smartctl unable to see the physical drive underneath. Sometimes
-d megaraid,Nor-d cciss,Nhelps; in other cases, it is easier to connect the drive directly for the receiving inspection. - A drive already installed in a NAS. Synology and QNAP systems include smartctl internally, but the installed version is often older than 7.4, and the web interface does not expose FARM at all.
For an incoming inspection, a direct SATA connection to a computer is therefore the most reliable option. What happens after you place the drive in a bay and create a volume is explained in how a NAS works internally.
If it is not a Seagate drive
FARM is a Seagate solution. It will not work on WD Red, WD Purple, Toshiba N300, or SSDs from other manufacturers. Even so, more options remain than you might expect:
smartctl -l devstat— Device Statistics is a standard ATA log rather than a brand-specific extension, so it is often available on WD and Toshiba drives as well.Power-on HoursandLogical Sectors Writtencan be read from it just as effectively.- SMART attributes 9, 12, 4, and, on some models, 241/242 (
Total_LBAs_WrittenandTotal_LBAs_Read). - The manufacturing date on the label, exact model and revision, and warranty status in the manufacturer’s system based on the serial number.
- Physical inspection: wear on the SATA connector, screw marks around the mounting holes, and dust in the seams. A drive that has spent years in a chassis will usually show it.
This evidence is weaker than comparing SMART with FARM, but it is still far better than nothing. The WD Red, Seagate IronWolf, and Toshiba N300 comparison examines how specific NAS product lines compare across brands.
Receiving multiple drives for a company NAS
For a shipment of four or eight drives, testing one and assuming the rest are identical is not enough. Check each unit separately—in the reported cases, a single shipment sometimes contained both legitimate and heavily used drives.
For every drive, save the model and serial number, a photograph of the label, the smartctl -x -l farm output, the hours from all three counters, the results of a short test followed by a long test (smartctl -t short, then smartctl -t long), the receipt date, and the invoice number. A long test on a 16 TB drive can easily take more than a day. For an array expected to last for years, that day is well spent.
The order also matters: checking the drive’s history does not replace a surface test, and a surface test does not replace checking its history. A drive with zero hours may have defective sectors, while a drive with a clean surface may already have spent five years in a data center.
If you are receiving a larger shipment, building a new array, or simply do not want to spend an evening poring over smartctl output, we provide incoming checks of drive models, operating history, and health as part of our IT management services for businesses—including a documented record for every drive in case of a warranty claim.
What to do when the figures do not match
Do not format the drive, add it to an array, or try to overwrite the values. You would lose evidence, and in the worst case the seller might interpret it as your own intervention.
Send the seller the model, serial number, invoice, photographs, and all saved reports. Describe the discrepancy in one specific sentence: the drive was sold as new, but the FARM log reports X operating hours and an assembly date of YYWW. A vague statement such as „I think it is used" is easy to dismiss; a number from the log is much harder to ignore. At the same time, verify the serial number in Seagate’s warranty system—a valid warranty is supporting information, not a substitute for a FARM check.
If you have already stored company data on the drive, the order is clear: first copy the data elsewhere, then deal with the seller. Securely erasing the drive before returning it should only happen after you have verified the data on another medium and saved the diagnostic reports somewhere other than the drive itself.
Frequently asked questions
Does FARM also work on WD Red or Toshiba N300 drives?
No. FARM is a log specific to supported Seagate drives. For other brands, try smartctl -l devstat and supplement it with SMART attributes, the manufacturing date, warranty status, and a physical inspection.
How many hours can a genuinely new drive have?
Zero is not a requirement because drives undergo testing at the factory. There is no single published limit covering every model and distribution channel, but single-digit hours are normal; thousands are not.
Can I perform this check in Windows?
Yes. Install smartmontools, open Command Prompt as an administrator, obtain the identifier using smartctl --scan-open, and then use it. SeaTools is Seagate’s graphical alternative, but a text report is better for a warranty claim.
Why does it not work through a USB enclosure?
The USB-to-SATA bridge may not pass the vendor-specific ATA commands required by FARM. Try another device type with -d, use a different bridge with ATA passthrough, or connect the drive directly via SATA.
Does a low number of hours mean the drive is healthy?
No. Hours describe the drive’s history, not the condition of its platters, heads, or sectors. A new drive can be defective, so after checking its provenance, always run a long test and review the SMART values.
Where to find us
ITHOPE is a company based in Brno, Czechia: we do data recovery from disks, RAID and NAS, PC and laptop service and IT for business, all in our own lab at Viniční 3587/216, 615 00 Brno.
Phone +420 775 556 063 (24/7 line). You can bring the device in person, or send it by courier (Zásilkovna, PPL).