How Reliable Is Seagate Hard Drives?

I hear Seagate hard drives internal or external are not that reliable and only last two or three years. Where western digital is more reliable.

There was web site not sure of name now where thay where saying %80 of hard drive will make it to 4 years. That means by the 4 years of service %20 of them while die.

It was study of hard drives used in cloud place.

Not sure of the web site name or if the information is false or not.

 

Backblaze is probably the site you saw.

Their data is always interesting but not sure it really applies to typical home users. Besides the obvious - we don't use drives in the same way data centers to - the drives we do use are not the same model numbers.

I've personally used drives from Seagate, WD and Toshiba and have not had one fail in more than 10 years. I always end up retiring them first.

And even if you buy the most reliable according to those statistics, you still need to keep current backups.

Having said that - SSDs are the way to go.

 

It would help if you could dig up the link for that web site so we could take a look at the report.

I don't see much about hard drive reliability nowadays since most people have switched to SSDs. However, Seagate hard drives seem to be as reliable as Western Digital's.

Having said that: all drives fail eventually, and usually unpredictably. Make regular backups of your data.

 

Pretty sure its the Backblaze site I linked too. I know of no other site that compiles such data. But again, their information may be useful for organizations using busy servers, but not likely for the rest of us.

My advise for buying hard drives is to avoid the budget models. For me, I look for the "Enterprise" lines. These drives typically are designed for 24/7 operation and come with a longer warranty. But they cost more too.

I should also note that nec209 asked about internal and external drives. There is a big difference in reliability between internal and external and that is because external drives have enclosures and most use USB to SATA adapters. Unless the enclosure has a fan, many of these enclosures trap too much heat. And if there is a fan, many are of poor quality and are noisy and/or fail prematurely. USB, despite its popularity, really is not a very reliable or robust interface. And the USB to SATA adapter found on many enclosures tend not to be very reliable either.

My point is, external drives introduce many more variables and potential points of failure compared to internal drives. So a lousy reliability statistic for an external drive does not necessarily mean the hard drive itself is unreliable, or that internal drives from the same maker will be unreliable.

And sorry, while SSD sales continue to climb as prices continue to fall, hard drive still outsell SSDs by a long shot. Most buyers still buy hard drives.

 

Hard drives will continue to sell until solid state price per GB comes way down. You get a lot more storage for your buck with a spinning hard drive right now. The last Seagate I had fail was mt fault, I some how stuck a desktop drive in my NAS and it lasted 2 years. The other thing to look at is who services the warranty, some drives you have to deal with the seller (usually OEM packed), some go straight back to the manufacture ( usually retail packed).

 

And they are getting there, but are not there yet. I only go with SSDs in my builds, but I don't have massive storage needs either.

One thing really driving the sales of SSDs is the mobile computer market. SSDs are lighter, smaller, consume less power for longer battery run times, and generate less heat - all desirable features in mobile devices.

 

Here is the link talking about it.

How long hard drives last.

https://www.prosofteng.com/blog/how-long-do-hard-drives-last/

 

Thanks for that link but please note the author based his entire report on the findings of the same Backblaze report I linked to above.

But more importantly, it should be noted Jeremy S., the author of that Prosoft rehash, failed to do his homework or verify his facts before publishing his article. Consequently, he presented some misinformation about the drives used. He claimed, "Backblaze used consumer-grade hard drives". That is false! Many, if not all of the drives analyzed were enterprise-grade drives.

A simple 5 minute Google search of the model numbers easily verifies this. For example, as seen here, the HGST HMS5C4040BLE640 is an "enterprise-grade" drive. And clearly, so is the Seagate ST100000NM0086 as is the Toshiba MG07ACAxxx series of drives.

So while it may be nice to know Company A makes reliable, heavy duty, "industrial" grade doohickies designed for the work-site, that does not automatically mean their light duty, "consumer" grade dookickies designed for the home are reliable too.

 

Honestly, 3-5 years has been the average lifespan depending on abuse, or use respectively when dealing with consumer mechanical drives.

As Digerati pointed out, some drives are built better than others.

 

A mech drive will last as long as it lasts!

If you dont back up any and all drives on your system, your a fool who is easily parted with his data!
No matter drive reliability

 

Thought this might be of intrest: https://hackaday.com/2019/12/10/patch-or-your-solid-state-drives-roll-over-and-die/
"YOUR DRIVE IS GOOD FOR ABOUT 3 YEARS AND 9 MONTHS"

 

This HDD is 10+ years old.

 

I can still access old IDE drives i've had since Jesus was a boy, although they are no longer in active service (backed up), I can still call them at will!

Love your drives with kindness, and they will love you back

 

That's good, but how a drive "ages" involves much more than just the passing of time. The number of hours the drive motor spins is certainly important, but if the R/W is just sitting there idle most of the time, then it is not being worked much. The operating environment during operation and during system downtime matters too. That is, operating temps, downtime temps, humidity and even the physical handling of the drive (less a factor with fixed PCs than with mobile notebooks).

That said, that report says the drive was powered on over 45,000 hours. 45,000/10/365 is over 12 hours per day. So that ATA/EIDE drive has certainly held up well!

 

One more thought: Avoid "refurbished" drives like the plague.

A refurbished drive is almost always a unit that was accepted back by the manufacturer under warranty. The manufacturer then (allegedly) tests it, reformats it and sticks a "re-certified" label on it.

The problem is the "testing" often misses issues that are likely to bite you in the a** when you attempt to install it or soon after. On my last go around with Western Digital, they sent me THREE DEFECTIVE "refurbished" DRIVES IN A ROW for a two year old WD Black I sent in under warranty (two failed a short DST; the third failed about 30 minutes into a long DST. The testing was done using Western Digital's "Data Lifeguard" software). I finally gave up and tossed the third one in the recycling pile.

If you have the option of buying an extended warranty that gives you an instant in-store exchange for a defective drive my advice is to go for it. The $10-$20 such a warranty costs is a bargain compared to dealing with incompetent manufacturer help desk "support" and the cost of sending off a drive only to receive a piece of crap "refurbished" one in return.

 

It would be nice if you could always avoid refurbished drives, but sometimes, you have no choice. As you correctly noted, if you send in a drive that is still under warranty via RMA, chances are you will get a "refurbished" drive. They don't give you an option.

The problem is, manufacturers and retailers cannot resell any opened box and/or returned product as "new". So they label them all as "Refurbished". But refurbished does not automatically mean the drive was faulty and they repaired it. In fact, in many cases, the drives are returned for any number of reasons that are not related to any fault. They were returned because they were the wrong size, wrong color or user error, or something else. Or perhaps there was a bad controller board they replaced and now it truly is refurbished and "good as new".

I will also say this and it applies to many products, not just drives. In many cases, it is not economically feasible for manufacturers to spend the resources (labor hours, training, parts, facilities, other logistics, etc.) to fix a returned product (unless an obvious, easy to fix problem). So those get tossed.

I would also say it is okay to blame the help desk support as incompetent, but take care not to blame the help desk support people as incompetent. Those first tier techs are poorly paid and more importantly, poorly trained or even receive no training at all, except how to follow a checklist. In fact, in many cases, if they deviate from the checklist because they recognize the problem, they get in trouble.

The problem with incompetent help desks is management failing to dedicate the necessary resources ($$$) to properly train and equip their help desk techs.

Over the years, we have had to return quite a few drives and I have to say, I never received a faulty one in return. You receiving 3 defective drives is very rare. I would say in your case, your stars were just way out of alignment. In fact, if you count the original drive, that's 4 bad ones in a row.

What really irritates me with some of those "refurbished" products is they often come with a 1-year warranty even though you may still have 2 or 3 years left on the original warranty. That seems like a rip-off to me and I suspect that could be challenged in the courts.

 

Is it just refurbished or just that all hard drives today are really tarrable today.

It seems hard drive capacity has increase so much today but performance gone way down.

I would take 200GB hard drive over 1TB hard drive any day.

The higher the capacity the lower the performance


The life hard drive of 3 to 5 years is outrageous.

 

Ummm, that's not true at all. Seek and access times have improved considerably over the years.

Do not confuse performance with reliability. They are two totally different things. Performance indicates how fast a drive can read and write data and today's drives are faster than those of the past. Part if that is due to the fact data density has increased significantly.

So even though rotation speeds have remained the same (7200RPM, for example), much more data is squeezed into the same amount of space on the disk platters. This means as 1 inch of a track zips by the R/W head, more data is read in on a 1TB drive than on a 200GB drive in the same amount of time. That yields faster performance.

Plus, today's drives typically have bigger and faster integrated buffers. That 200GB drive might only have an 8MB buffer while a 1TB drive might have a 32MB or even a 64MB buffer - or even a big chunk of flash (SSD type) memory for its buffer. A larger buffer means the OS can hand off writes to the drive much faster and more data can be read into the buffer. All relate to better performance.

Reliability (drive failure or malfunction) is something totally different. And while I cannot find any specs on that, I suspect that has gotten worse for a couple reasons that have nothing to do with the drive's capacity. First, drive makers are constantly seeking ways to cut production costs and increase profits. And sadly, that often means cutting corners, taking shortcuts, and using cheaper parts - at least on their cheaper models. You can often see this reflected in their warranties. In years past, most drives had a 3 year warranty with a 5 year warranty on their Enterprise-class drives. You can still find 5 year warranties on the better drives, but 1 year warranties are more common than ever.

Another problem I see is more and more hard drives are going in cheap USB enclosures. The USB interface is not always as reliable as it should be. The USB to SATA adapter inside the enclosure can fail, as can the enclosure's power supply. Any failure in those areas are often considered as "drive failures" even though the actual drive inside might still be working fine. That can skew the failure numbers.

All drives will fail - eventually. And hard drives being electro-mechanical devices are likely to fail sooner than devices without moving parts. So as always, make sure you have a viable backup plan.

IMO, this all just means it is time to migrate to SSDs.

Merry Christmas!

 

I suspect many "performance" issues on higher capacity HDDs where the operating system (Windows) resides are caused by not defragmenting the drive often enough.

A drive is "fragmented" when part of a file is stored in one area and other parts are stored in other parts of the drive. If a drive gets 0.25%-0.50% fragmented it starts to slow down. At 2%-3% running a large program or accessing files can be like watching paint dry.

It goes without saying that a 1TB drive with 500GB of files will naturally have far more file fragments than an old 80GB drive that's half full.

Like Digerati said, solid state drives are the way to go. You can get a 240GB Sandisk for around $40 at Best Buy or Amazon (most people have 100GB or less stored on their drives, even with thousands of photos and songs on it). Even a 1TB Samsung SSD is now hovering around $100.

 

Umm, sorry but that is just not a problem anymore - and certainly not at percentages that small. I mean .25 percent? I don't think there is any defragging program that can defrag a drive that efficiently. Even 2 - 3% represents a very small amount. Do you have any links to support those numbers?

Actually, that is not true. You actually have it backwards. The larger the drive, the more space there is to ensure all the file segments can be saved together. And the larger the drive, the more likely there will be more free disk space on the drive - perhaps significantly more than that 40GB you mention for the 80GB drive. That again gives the larger drive lots more space to ensure complete files remain intact (not fragmented).

As long as the drive (regardless its size) is not running critically low on free disk space (another reason to buy monster drives), fragmentation will not in any significant way affect performance UNLESS the user negligently changed the Windows default settings!

And to that point, it is critical to note since Windows 7 (maybe even in Vista), Windows defrags complete hard drives automatically once a week. It does this way in the back ground so it does not affect performance while the user is actively performing tasks. Once a week ensures that is more than enough - even if defragging is cut short by the user turning of the computer after every session. So as long as the user did not intentionally disable defragging (and there is no reason to do that), hard drives are regularly being defragged. Note even XP automatically defragged boot files.

The ONLY reason a user should disable that feature in modern (W7 and newer) Windows is if they install a 3rd party defragger. But that is not necessary. Windows' own is more than adequate.

Also, because today's very large hard drives have much faster seek times (the time it takes to find the first file segment on the disk) than older drives, and because these drives are so big, it is no longer necessary to bunch all the files at the front of the disk. So it does not matter if files are scattered all over a disk as long as the file segments are not scattered about (which is what auto-defragging mitigates).

Plus, because virtually all computers are, or are expected to be, connected to the Internet, operating system and program files are regularly being modified/updated. So when Internet connections became the norm, fragmentation actually became a bigger problem when defrag programs consolidated free disk space by jamming all program and data files at the beginning of the disk.

Before some complain about the defragging feature of Windows Optimize Disk, it is true some 3rd party disk defrag programs are more efficient than Microsoft's. But that is immaterial (contrary to the marketing "hype" of those 3rd party apps). This is because even after the most efficient defragging, fragmentation starts all over again as soon as the computer is used after defragging. This is because immediately, files are opened, temporary files are saved, files are modified then saved in a different location. Then the old version is deleted leaving a hole behind, spots where new file fragments are stuffed. Its an endless cycle - thus the reason Windows defrags automatically once a week. So yes, a 3rd party defragger may provide a more efficient defrag, but any (if any) advantage that provides is quickly leveled out when the computer is put to normal use again.

So the user really has no responsibility here other than to ensure they don't run out of free disk space - which has nothing to do with how big the disk is.

 

This is why I always partition 1TB hard drive into three parts. If it was 300GB hard drive I would not partition it. If it was 4TB hard drive I would make more partitions.

Windows defrag does not work well with large hard drive and modern hard drives. That is why everyone say to use third party defrag.

I'm not sure how often in a month you defrag your hard drive.

 

Where did you read this? Please provide the links.

People recommend third-party defrag programs because they are faster and have more options (like optimal placement of frequently accessed files and elimination of gaps that encourage future fragmentation). Windows' built-in defragger doesn't have any "bells or whistles", but it does perform a decent, basic defrag of your files.

 

Sorry, but that is totally wrong! On both counts.

Let's start with the second part first. By far, most users, including professionals and enthusiasts do NOT use or recommend using alternative solutions anymore. Why? Because contrary to the first part of your comment, the defragging tool native to Windows works just fine, even on monster drives. In fact, it (as do all defragging programs) works particularly well on monster drives because it has so much room to work in.

Sorry but again your reasoning makes no technical sense at all. Partitioning does absolutely nothing to improve performance of the drive, or the defragging utility. If anything, it can degrade the defragging function because it limits the amount of free space in each partition.

The only reason to partition a drive is for "user convenience" for those who use File Manager to manage their program locations. If you manually moved a file to a secondary partition, it might be easier to manually find it through File Manager. But with modern Windows, Libraries do that so there is no need (and for most users, no advantage) to partitions. In fact, partitioning can force the read/write head to jump around more. And of course, more complex partition tables introduce more potentials for disk corruption.

It is time to stop treating modern versions of Windows like we did XP. Managing files on monster drives can just as easily be done in a single partition using folders.

I'm curious. Did you read my post that you quoted? Windows own defragging tool defrags automatically once a week. There is no need to manually defrag a hard drive.

While perhaps some of that is true, these really are not valid reasons any more. Let me address each point (though I already did above).

1. Faster - True some 3rd party defraggers are faster at manually defragging. But so what? Since Windows defrags automatically every week unnoticed in the background when the user is idle, why does speed matter? There is no need to manually defrag. If on some rare occasion you do want to defrag, it should be so rare, again it should not matter. I mean it is not necessary to sit there and watch a drive being defragged. I'd rather watch paint dry.

2. Elimination of gaps - Not a factor anymore because (1) as noted, computers are connected to the Internet so Windows and installed programs are frequently being updated with modified/updated files. Only "static" files like songs and videos rarely (if ever) change and so they don't become fragmented again (if they ever were). And (2) with today's larger drives (as long as the user did not create a bunch of smaller partitions) there should be plenty of free disk space just to avoid such gaps that might "encourage" fragmentation. However, Windows defragger still does a lot of free space consolidation just for that reason too - but without jamming all the used space together which actually promotes fragmentation.

3. Bells and whistles just aren't needed. A basic defrag program is all that is needed. Those bells and whistles are really just marketing "fluff" because, again, as soon as users start using their computers again files will be moved about as they are opened, modified and closed again. You all are aware of the KISS principle, right? Keep It Simple Stupid works best for defragging programs.

4. Disk space - 3rd party defraggers take up disk space unnecessarily (okay, this is minor but still a valid point).

5. Optimal placement. This is actually a BIGGIE! As I also noted above, seek times on modern hard drives make such optimal placement unnecessary - in terms of putting frequently used files at the beginning of the disk. Years ago, jamming all your most used files at the beginning of the disk helped them load faster but that is no longer true because modern drives are much faster and have large on-board buffers. Not to mention Windows prefetch routines learn user patterns to preload those programs into the Page File for even better load times (unless the user dinked with those settings too! ).

My best advice is to let Windows manage defragging and your page files too. It knows how to do both very well.

 

Prefetch loads all programs into Cached memory, temp/installers, everything, based on total history, not usage. The Page File is mainly used to store Edit history for Notepad and MSOffice programs and as a memory pressure overflow from Active memory.

To limit the cache preload to the extant files and those that you use regularly, Use CCleaner to remove old Prefetch entries first, then load up all your usual/regular software, run RAMMap and use it to check the loaded files with File Summary - pay attention to the extent of the temp/installer files preloaded - then run a disk cleanup and clean all inactive (ie, not from software loaded currently) temp folders. Then use RAMMap to clear out all Cached memory data: Empty > Standby List - nothing else except File > Exit.

To prevent the defrag routine's optimal placement wasting more time disk thrashing, you could manually edit the C:\Windows\Prefetch\Layout.ini - but it'll be a very long and boring task to find/remove all invalid entries, best not to bother.

Defragging works best with nothing else running and a very large chunk of free drive space, switching to a 3rd party defragger might lead to excessive drive thrashing on the first run if it uses a different algorithm to the MS defrag routine. It gets worse if you try comparing multiple defraggers, that excessive drive thrashing could happen every time.

 

There are various fetch and cache routines that Windows uses but it is important to note several are disabled when Windows detects it is running off a SSD.

 

I disabled my own as MS' blurb can't be trusted. Check RAMMap and the prefetch data for yourself.

 

That me use a example why I partition. Think of small ant on a dinner plate if you make three section on the dinner plate the ant has less surface area to travel. If there is large plate and no section you have more surface area.

Other thing is fragmentation. An one large surface area can have fragmentation any where. Where the partition will contain the fragmentation to its area.

 

So you use windows defrag and have it turn on? And are happy with it?

 

That's not a valid argument - especially these days. It seems you did not follow what I was saying above about modern drives being so much faster and having much bigger buffers.

To begin with, it is the drive's "seek time" that matters here. That is, how long it takes for the read-write head to find the "first" segment of the file. 25 years ago hard drives were much slower, seek and read times were several milliseconds slower, and buffers were much smaller too. Back then there was some validity to that argument. But not today.

If you research Seagate's entry level 500GB drive, you will see it has an average read seek time just 8.5 milliseconds! That's incredibly fast - twice as fast and those older drives. To put that in perspective, for the average human to detect a "flicker" in the lights, that flicker must have a duration of around 30ms. This is the primary reason old analog movies, that were filmed at 24fps (one frame every 41.66ms) did not "flicker".

That drive also has a 32MB buffer. Most drives 20+ years ago had buffers of just 2MB. Expensive drives had only 8MB. A bigger buffer results in much faster reads, and it allows the OS to hand off writes to the drive much faster too.

So even if you were right and the seek time for that first file segment is longer, (1) it is only by a couple imperceptible milliseconds and (2) the loading of the remaining file segments would still be much faster more than making up for any seek time differences.

Also, consider where the landing zone is on hard drives. As seen here, it typically is near the center. When idle or when powered off, the read/write head "parks" in the landing zone. So using your plate analogy, it makes no difference in time which section of the plate the head must move to to find (seek out) that first file segment.

No. Sorry. Also invalid. In fact, it is backwards logic. With one big partition, the file system can use the entire drive to help ensure the entire file is save whole and unfragmented.

Fragmentation is minimized when there is lots of free disk space available. When you partition you divide that one large shared chunk of free disk space into smaller, un-shared chunks. That contributes to fragmentation - not good.

This takes us back to the seek times. With 1 big partition, once the first segment is found, because the file is still whole, the entire file is then read sequentially into the buffer much faster without the R/W head having to jump about.

And again, unless users dinked with the defaults, Windows will keep fragmentation at bay by defragging automatically every week. So fragmentation NEVER becomes an issue - UNLESS running critically low on free disk space - which is 3 times more likely when there are 3 partitions instead of 1!

Partitioning typically degrades performance - not improve it. This is especially true if you put your OS on one partition, your programs/applications on another and your data on a third. This forces your system to constantly access the partition tables (slowing access down) then it causes the R/W head to jump across multiple partitions as it accesses an OS file, then a program file, then a data file.

I say again, the ONLY reason to partition a drive today is to make organizing folders and finding files easier for those users who rely on File Manager. There is no performance advantage.

If you want to improve performance, but want to take advantage of using 3 different drive letters, use 3 separate drives! Don't partition 1 drive. With 3 drives, the OS can access data from all 3 drives at once. That's not possible with partitions because the single R/W head cannot be in 3 places at once.

Of course the best solution is to migrate to all SSD. Then fragmentation is never a problem.

So do you - unless you manually went in and intentionally disabled it.

 

Yes, it is turned on. However, my PC no longer has a hard drive in it, and Windows is smart enough not to try to defrag my SSD (it runs TRIM instead). Back when I had hard drives, I left the Windows defrag on; never had any problems with it.