Recent report claims Seagate and Hitachi HDDs are most likely to fail

[AlphaX]

In context: While the popularity of SSDs continues to increase, some users still prefer to have a trusted hard drive for their affordable costs and high storage capacities. However, a recent study shows buyers may not always pick the most reliable HDDs.

Recently, the typical usage of a hard disk drive (HDD) has shifted drastically. As solid-state drives (SSDs) have become the primary storage device, HDDs are the go-to for games, file backups, or cheap additional storage. Hard drive prices have dipped significantly in the last few years, leading some buyers to include one in their computers just in case. So finding a reliable one is just as important.

Timothy Burlee from Secure Data Recovery wanted to find out why HDDs fail and which manufacturers are the most reliable. So he tested 2,007 broken HDDs from six manufacturers--Western Digital, Seagate, Hitachi, Toshiba, Samsung, and Maxtor. Capacities ranged from 40 GB to 10 TB. He obtained the drives from customers in 2022. Berlee recorded each HDD's power-on hours and pending sector counts to determine their longevity and fail conditions.

Toshiba came out on top with an average lifespan of 34,799 hours. Hitachi finished dead last, managing only 18,632 hours. Maxtor performed impressively, nearly reaching 30,000 hours. Western Digital and Seagate HDDs clocked in around the mid-20,000-hour ranges, while Samsung scored at just over 19,000 hours, barely better than Hitachi.

The next metric Burlee looked at was "pending sector counts." Pending sectors are bad subdivisions of the discs. The average count for all drives studied is 1,548 sectors. Burlee notes that a hard drive with 1 TB of storage has "just under 2 billion total sectors." While the average may seem small, bad sectors will rapidly accumulate until the HDD becomes unusable. So higher counts are an indicator that the device is headed for failure.

In this test, Hitachi finishes last again, with an average pending sector count of 3,348—over twice the average. Seagate followed close behind at 2,671, and Toshiba didn't perform very well either, averaging almost 1,900 pending sectors. Western Digital, Samsung, and Maxtor faired far better, scoring well below the average with 628, 529, and 228 bad sectors, respectively.

Burlee estimated that of all failed HDDs he received, 80 percent suffered non-predictable failures, which include electrical issues, malware, natural disasters, or simple mishandling. He excluded these drives from the study since he was looking for HDDs that failed from regular, predictable use. It's still interesting that most hard drives fail from external factors.

Burlee concludes that the test shows that a drive slated for failure starts showing signs after about 25,233 hours or two years and 10 months of use. If your hard drive makes it past that time frame, it should be able to survive much longer. Interestingly (and perhaps unsurprisingly), many manufacturers only have 2-year warranties on their consumer hard drives, so be cautious.

Permalink to story.

https://www.techspot.com/news/97990-recent-report-claims-seagate-hitachi-hdds-most-likely.html

 

[redgarl]

Surprising, because Toshiba Enterprise drives are far from the best. I have four 16 TB drives in my NAS and some are making noise like a small hum when functioning. The quality is just not there. However, you can buy them for as low as 250$ if they are on sale.

As for reliability, Blackblaze report Toshiba to be between WD and Seagate when it comes to failures.

 

[sdms96825]

So, looking at the power-on hours and current pending sector count, it appears that Maxtor is the most reliable.

However, only 27 Maxtor disks were considered compared to 936 for Western Digital.

Is the sample size for Maxtor too small, or is it just because not enough Maxtor drives have been failing?

 

[yRaz]

So, looking at the power-on hours and current pending sector count, it appears that Maxtor is the most reliable.

However, only 27 Maxtor disks were considered compared to 936 for Western Digital.

Is the sample size for Maxtor too small, or is it just because not enough Maxtor drives have been failing?

I don't know how I feel about it, I do think the sample size for Maxtor is too small but their reliability is high enough in that sample that I wouldn't ignore it.

 

[noel24]

I'm confused. Weren't Hitachi, Maxtor and Samsung HDD departments swallowed by WD and Seagate respectively?
How old are the drives in this study?

 

[DaveBG]

This is so subjective. So far many reports showed that Hitachi are the most reliable, Seagate was improving too and here we see the exact opposite. While I have early Samsung drive that is rock solid after a lot of trime

 

[kiwigraeme]

This is so subjective. So far many reports showed that Hitachi are the most reliable, Seagate was improving too and here we see the exact opposite. While I have early Samsung drive that is rock solid after a lot of trime

Again without digging down deep - kind of hard to know -old 4Tb drive and new 16Tb drive from same company may be a silly comparison

Plus source of drives etc - you read on Amazon _- I bought 8 drives for my NAS 6 failed ( TBF he removed this type of failure )

So some batches are good - some bad - was it from Thailand or from China etc

Plus removing other failures takes away a huge source of failures

Failing disc drives is fairly good as time to recover= but other sources ( with out spending money new controller etc - then it's gone )

Was audit the audit - did he test the accuracy of these reports - probably not - as imagine a test would take a long time

 

[truthmonger]

I knew Hitachi were sub-standard but Seagate kind of surprises me. I thought their bad days were behind them. Toshiba were always the worst.

 

[NintPlayBox]

Am I back in 2004?
Or this is very old data, because I haven't heard of a Hitachi or Maxtor HDD since back then.

 

[torchyear]

The devil is always in the details in these kinds of reports. Making assumptions based on brand alone is disingenuous. Backblaze reported a single HGST 8TB model (and only 4 samples) as the outlier for otherwise low failure rates in this brand during 2022. The lifetime AFR stats for Seagate somewhat correlate with this report from SDR, but once again it's dependent upon specific SKUs.
Edit: also the SDR methodology was to sample already-failed drives, and not measure incidence of drive failure in a functioning sample over time. Logged usage in failed drives doesn't necessarily correlate with an unreliable brand - that's a somewhat logical leap of faith.

 

[Athlonite]

I have disks that are 70510 hours or more (thats a tad over 8 years) that are still ticking along like they were still brand new it all comes down to optimising their running conditions like keeping them cool (50c or less) and vibration free taking them out and dusting them off once or twice a year and careful handling when they're out of the case

 

[daffy duck]

My brand new Toshiba X300 8TB was making scrapping sounds the moment I installed it, had to be replaced. The the only time I have a HDD DOA. In total I had 2 hdd's fail, one WD, one Seagate. But I have several hdd's over 7 years old still going strong and a few over 10 years.

Can you still buy Maxtor? I don't recall seeing that name in what seems like forever.

 

[anthonymaw]

Hard drives, both HDDs and SSDs are like a mine field: You just never know when it will blow up and evaporate all your data with it. It cannot be emphasized enough to BACKUP all your data frequently and reliably. With the advent of these modern multiple terabyte hard drives, the user stands to LOSE more data than ever. Sometimes they fail gradually - the performance slows down over time until you get fed up with waiting for that hourglass or spinning donut like at boot time or launching programs but also as mentioned in the article there are also catastrophic failures involving the controller boards, heads and platters. Thermal cycling seems to be a factor sometimes as solder joints heat up and cool down and can develop micro fractures on circuit leads. A data recovery job can cost $1000 to $10000.

 

[yeeeeman]

I've had many seagates fail...in pcs, laptops. western digital, never. also have an old hitachi from 2000s and one for laptop from 2008, still running good. maybe now with the buyout of hitachi (wd), they are worse quality.

 

[KofeViR]

My experience with Seagate has been terrible. Almost on Maxtor-level (and Seagate bought Maxtor). I avoid Seagate when I can.

 

[arrowflash]

My own experience is that desktop Seagate HDDs really are ticking timebombs, and WD HDDs are a lot more reliable.

External Seagate HDDs, however, tend to be very reliable and durable, both the 3.5" ones with external power supplies and the portable varieties. My oldest external Seagate HDD is nearly 13 years old at this point and is still going strong.

 

[AIC1Drew]

RAID

 

[toooooot]

No. HDDs used to be reliable. Now they are only suable in data centers. What is the point buying HDD for personal use if it can likely fail in 3-5 years?
No surprise fewer and fewer people rely on them.

 

[daffy duck]

Not all Segates are bad, backblaze has the data to show that. You just have to be very careful about which model you buy. Check the model number very carefully.

 

[leonvh]

In my professional experience I have been able to determine that the hard disks that have presented failures and fatal errors during the last 15 years, with great difficulty to recover the data, have been mainly Samsung and Hitachi. The ones that have shown the best performance are Western Digital. The hard disks evaluated were mainly 500 GB and 1 TB capacity, we have not used the Seagate brand and therefore we have no reference of this brand to consider in our experience.