Seagate finally offering a down to earth priced internal 4TB HDD.

can i ask how you burn in the drives?

i am working on building a test bed for server parts to test before going into production.
 
can i ask how you burn in the drives?

i am working on building a test bed for server parts to test before going into production.

partedMagic boot CD/USB and then run short/long SMART tests followed by say 24hrs of badblocks
 
I even burn in "enterprise" drives. 30+ hours of bad blocks before I put a single byte of important data on a drive.

I have seen quite a few enterprise SAS drive failures in a few Dell servers that we use. These are ones that I do not provide maintenance so I do not know what drive model or even manufacturer has failed but I do see the failures..

Agreed with both, though I mentioned consumer drives specifically since that's what we were discussing.

Having seen a fair number of enterprise SAS drives fail myself, I'm definitely not one to believe that enterprise drives are a magic fix for reliability. :p

As for burn-in: I know badblocks is what folks like to prefer but I personally have ran a couple on-off cycles of DBAN when I didn't have badblocks handy.
 
I have seen quite a few enterprise SAS drive failures in a few Dell servers that we use. These are ones that I do not provide maintenance so I do not know what drive model or even manufacturer has failed but I do see the failures..

Interesting...I'd be interested in knowing what percentage of failures you are experiencing (failures vs total drives). Enterprise drives can and do fail, but on a whole, I trust them much more than the typical consumer-grade drive. It's not just the quality of the drives, but the validation their firmware is given and tolerance they have to things like vibration. There is still a chance of an issue, just less of chance...;)

The crappy thing about any hard drive is that you have no idea how it was handled en route prior to the end user taking possession of it, so that aspect throws a wildcard factor into the mix. A poorly handled drive is prone to fail...regardless of what category it's in.;)
 
Interesting...I'd be interested in knowing what percentage of failures you are experiencing (failures vs total drives). Enterprise drives can and do fail, but on a whole, I trust them much more than the typical consumer-grade drive. It's not just the quality of the drives, but the validation their firmware is given and tolerance they have to things like vibration. There is still a chance of an issue, just less of chance...;)

No offense but you base "less chance" of SAS/Enterprise disks failing than SATA, on a feeling?

SATA disks are fine for RAID, and no large scale tests or studies I'm aware of prove failure rates for one over another. Obviously for enterprise and when youre spending other people's money you go SAS, but for purposes of this forum it hardly matters since most people here are just storing media and movies at home and building home media servers.

SAS drives are a waste for that usage scenario, since for the same price as a SAS disk you can buy two SATA of equal size and not only have a 1:1 physical backup but money left over.
 
The risk is mitigated if you design your system around the fact that they're consumer-grade drives. I also believe that burn-in testing is mandatory for consumer drives, it weeds out a lot of early deaths.

No offense but you base "less chance" of SAS/Enterprise disks failing than SATA, on a feeling?

SATA disks are fine for RAID, and no large scale tests or studies I'm aware of prove failure rates for one over another. Obviously for enterprise and when youre spending other people's money you go SAS, but for purposes of this forum it hardly matters since most people here are just storing media and movies at home and building home media servers.

SAS drives are a waste for that usage scenario, since for the same price as a SAS disk you can buy two SATA of equal size and not only have a 1:1 physical backup but money left over.

I wasn't advocating SAS over SATA and I completely agree, for most intents and purposes, SAS is definitely not required. I was comparing SATA consumer grade drives to SATA Enterprise Nearline drives, which carry a longer warranty and in my experience are more reliable for what sometimes is only a modest premium over the consumer drive.

Personally, I bought the SAS drives that I currently use so that I have the ability to reuse my drives in a large external storage chassis using the drives' dual-port feature with dual RAID adapters to provide better redundancy in the future, but I would certainly not advocate that strategy for most people, as it's entirely unnecessary.
 
In my experience the enterprise drives are not just a "modest premium" over the consumer version.

Seagate Barracuda 7200.14 ST3000DM001 3TB 7200 - $140
Seagate Constellation CS ST3000NC002 3TB 7200 - $240
 
I personally dislike seagate drives, I had a multiple failures of their laptop drives and I am never going to buy them again...
 
In my experience the enterprise drives are not just a "modest premium" over the consumer version.

Seagate Barracuda 7200.14 ST3000DM001 3TB 7200 - $140
Seagate Constellation CS ST3000NC002 3TB 7200 - $240

You're probably right most of the time, but sometimes the difference is more in the range of $50 or less.

I compared the WD 2TB Black to the WD Enterprise 2TB online and found roughly a $50 premium for the Enterprise drive. Given the feature set of the Enterprise drive, the premium is worth it (at least to me) to ensure reliability in a RAID array, especially one with a significant number of drives (and the increased vibration that comes as a result).

Depending on the amount of space needed, a smaller array with higher quality drives may be a better idea than a slightly larger array with lesser drives, to assist in mitigating some of the increased cost that comes with the Enterprise drives. Ultimately, it's up to the person building the array to weigh the array's characteristics and their priorities and decide for themselves whether or not the drives they've selected are the best choice...;)
 
I have significant distaste for consumer-grade hard drives, especially when the size is now up to 4TB. It's a lot of data to risk losing to a lower quality drive. I firmly believe that getting the Enterprise Grade version (Constellation ES.3), while being a fair bit more expensive is the safer way to go, as they're built to a much higher standard.

It all boils down to how much you value the data contained on the drive, I guess...;)

You will never see a serious enterprise risking data on any kind of drive. They will always have backups. You will however see enterprises using consumer drives. So, you're doing things wrong. Spend your money on backups.
 
So, you're doing things wrong.

I disagree. I've used a total of 14 Seagate Constellation ES 1TB drives, a full case (20) of WD2002FYPS enterprise 2TB drives and my latest array, the 12x Constellation ES.2 3TB SAS drives in my sig, which has run non stop now for nearly two years. The other arrays also ran 24x7 for years without issue.

So on the contrary...I'm doing something very right...

I do have things backed up, but consider it's use an absolute last resort. I'd much rather not have to deal with failures at all, which is why I use the drives I do.
 
How many consumer drives have you used? From what I've seen odditory alone has had quite a number of consumer drives in raid without issues and I've used a fair amount myself at home. Just because you haven't had a failure does not mean you're "doing something very right".

I'd much rather not have to deal with failures at all, which is why I use the drives I do.

Heh, I've seen enough failures with enterprise (both nearline and full spec) drives that I find that attitude a bit funny. :p


As a side note, it seems the egg is still occasionally ships drives poorly:

if4jky8l.jpg


Previous 5 drives came via full 20 pack OEM crate, this time around it had the drives essentially exposed to one side of the box. Couple of the bubbles were popped so obviously they had impacted the side at least a few times. I think I might just not bother with a burn-in test and just RMA it, but I'm not sure the next shipment will be any better.
 
How many consumer drives have you used? From what I've seen odditory alone has had quite a number of consumer drives in raid without issues and I've used a fair amount myself at home. Just because you haven't had a failure does not mean you're "doing something very right".



Heh, I've seen enough failures with enterprise (both nearline and full spec) drives that I find that attitude a bit funny. :p


As a side note, it seems the egg is still occasionally ships drives poorly:

if4jky8l.jpg


Previous 5 drives came via full 20 pack OEM crate, this time around it had the drives essentially exposed to one side of the box. Couple of the bubbles were popped so obviously they had impacted the side at least a few times. I think I might just not bother with a burn-in test and just RMA it, but I'm not sure the next shipment will be any better.

I used probably about a dozen over the years before using the drives I mentioned above. I never ran them in a RAID array though. I'm not saying that enterprise drives are immune to failure, it's just the fact that they're validated to a higher level than the consumer grade drives are and built to a higher standard to last their longer warranty period. Case in point...the Barracuda 7200.11 1TB drives had a significant firmware issue a few years back...the Constellation ES drives had no such issue, even though the drives were very similar. I think the eventual fix involved Seagate utilizing the Constellation firmware to fix the issues with the Barracuda. Avoiding that issue for me was more than worth the premium in price that I paid for the Constellation ES drives.

Consumer drives may indeed work with no issues and may be an effective offline backup. But the odds are better that the enterprise drives, which are validated for 24x7 operation in a RAID array, will perform reliably over a longer period of time. If folks manage to get large numbers of consumer drives working in a large RAID array...great. I'd personally prefer to spend the extra money to avoid the potential headaches. I don't really care if some people think I'm "doing things wrong" by using enterprise drives. The fact that I've had no headaches or failures is worth the extra money to me and I'd prefer not to roll the dice on a consumer-based solution that has the potential to waste a great deal of my time and increase the risk of data loss due to some stupid issue. I don't have a vast excess of time to spend screwing around with hardware issues on my main machine and even if I did, I have no such inclination. As I said before though, I may consider consumer drives for an offline solution, but only after careful analysis of the drive in question and after it's had a chance to mature a bit and have any bugs ironed out.

One can never control how drives are packaged by retailers. It's the one wildcard that no end user has control over. Unfortunately, there is not much we can do about it, except by ordering full manufacturer sealed cases and even then, you're at the mercy of the courier of shipping service.
 
My main point of contention are your statements along these lines:

I'm not saying that enterprise drives are immune to failure, it's just the fact that they're validated to a higher level than the consumer grade drives are and built to a higher standard to last their longer warranty period.

But the odds are better that the enterprise drives, which are validated for 24x7 operation in a RAID array, will perform reliably over a longer period of time.

I'm not saying you're doing things wrong, but in my own experience as long as drives are burn-in tested and the early death drives removed, there isn't a significant difference of drive failure rates between enterprise, nearline, and consumer drives in the arrays I've worked with. I completely understand what the marketing stuff says, but I have not seen any reliable study that has shown what you say is the case.

Btw, the Barracuda ES also had the same issue that the 7200.11 had: http://web.archive.org/web/20120214...agate/crm/selfservice/search.jsp?DocId=207963
Perhaps you simply got lucky and had a model that was not affected, but from that KB article you can clearly see that some models of the Barracuda ES line had the exact same issues as the 7200.11 even though they were "validated to a higher level".

You like your enterprise drives and think that it's foolish to use consumer drives in a raid. I like saving my money and using consumer drives and think that there isn't much difference in reliability between consumer and enterprise drives (as long as you burn-in test). We'll just have to agree to disagree :)
 
Last edited:
My main point of contention are your statements along these lines:



I'm not saying you're doing things wrong, but in my own experience as long as drives are burn-in tested and the early death drives removed, there isn't a significant difference of drive failure rates between enterprise, nearline, and consumer drives in the arrays I've worked with. I completely understand what the marketing stuff says, but I have not seen any reliable study that has shown what you say is the case.

Btw, the Constellation ES also had the same issue that the 7200.11 had: http://web.archive.org/web/20120214...agate/crm/selfservice/search.jsp?DocId=207963
Perhaps you simply got lucky and had a model that was not affected, but from that KB article you can clearly see that some models of the Constellation ES line had the exact same issues as the 7200.11 even though they were "validated to a higher level".

You may be entirely correct about there being no difference in the drives after a burn-in, and indeed the drives may be very similar overall and even may be relabelled nearline drives with 99.9% identical firmware. We've seen it from manufacturers many times before. I'm just less willing to take the chance of an issue, considering the data and time at stake. With a less important portion of my system, in which correcting a failure is a merely a matter of a component replacement, I'm entirely willing to do play around and explore that theory. Indeed for a backup array, it's something that may be definitely worth considering. But for my primary array, the risk is too great for me to take.

I took a look at that link you posted...took me a sec to see it, but it references the Barracuda ES drives. To the best of my knowledge, the Constellation ES drives were not afftected by the issue and served as the basis for the eventual fix.
 
I took a look at that link you posted...took me a sec to see it, but it references the Barracuda ES drives. To the best of my knowledge, the Constellation ES drives were not afftected by the issue and served as the basis for the eventual fix.

Do you have a source for that statement that the Constellation ES firmware was used as a basis for the eventual fix? I did a bit of research and it looks like the Constellation ES series actually replaced the Barracuda ES series, so it's highly unlikely the Constellation ES firmware was used to fix the Barracuda:

Seagate press announcement of the Constellation series, dated Feb 02, 2009

It seems Seagate doesn't put dates on their KB articles, but The Tech Report has a news article announcing the Barracuda/Barracuda ES fix for the 7200.11, dated January 22, 2009

Regardless of that, the Barracuda ES drives are enterprise drives so my original point still stands.
 
Do you have a source for that statement that the Constellation ES firmware was used as a basis for the eventual fix? I did a bit of research and it looks like the Constellation ES series actually replaced the Barracuda ES series, so it's highly unlikely the Constellation ES firmware was used to fix the Barracuda:

Seagate press announcement of the Constellation series, dated Feb 02, 2009

It seems Seagate doesn't put dates on their KB articles, but The Tech Report has a news article announcing the Barracuda/Barracuda ES fix for the 7200.11, dated January 22, 2009

Regardless of that, the Barracuda ES drives are enterprise drives so my original point still stands.

The Barracuda ES issue you linked to isn't the infamous 7200.11 drive detection issue. You're right...the Constellation ES did replace the Barracuda ES. The Barracuda ES came long before either the 7200.11 or the Constellation ES. I don't see the point in that link...you're proving that enterprise drives have also had issues? Sure they have, but I'd wager fewer than consumer drives have had.

I don't recall where I read that the Constellation ES served as the basis for the fix, but I recall reading it years ago. The fact remains that the Constellation ES never experienced the issue that plagued the 7200.11. I'll try to find where I read that though.

I don't see anything in the Tech Report article referencing the Barracuda ES, only the Barracuda 7200.11.
 
http://web.archive.org/web/20111013...elfservice/search.jsp?DocId=207963&NewLang=en

A firmware issue has been identified that affects a small number of Seagate Barracuda 7200.11 hard drive models which may result in data becoming inaccessible after a power-off/on operation. The affected products are Barracuda 7200.11, Barracuda ES.2 SATA, and DiamondMax 22.

Oh well, we've spent enough time on this OT...back to our regularly scheduled consumer level Seagate 4TB drive goodness.
 
Last edited:
The ST4000DM000's specs are not properly listed. The STBD4000400 is a 5900 RPM drive , it has 64MB of DDR2 and ERC so its perfectly fine for NAS setups or any kind of back up solution.

The site that shall not be named is currently selling them with tons in stock (of the STBD4000400) for $204. So if you want a solid NAS drive that's bigger than 3TB's this might be a pretty hot drive in the coming months.
This is getting interesting now. I asked my supplier of the disk about the specs for the ST4000DM000 and they confirmed that it is 7200 RPM with 2 years warranty. They said that this information came directly from Seagate.

It will be very interesting to examine the disks when I get them, that's for sure.
 
I have the ST4000DM000's and under "device info" in partedmagic it's reporting 5900rpm. As a side note I think two of my drives might be duds due to newegg fail shipping. They're going through the badblocks process at nearly half the speed of the others though badblocks itself hasn't reported any bad sectors yet.
 
Last edited:
The disks have been shipped! Hopefully I will know exactly what they are tomorrow. I already know the serial numbers and I can confirm that the warranty is 2 years (Expiration 24-May-2015) by checking it on Seagate's website. It gives the same expiration date for both the US and Sweden.

Any hints on how to run a proper bad blocks scan?
 
Someone needs to record the sound of the drive and frequency analyze it !
 
Someone needs to record the sound of the drive and frequency analyze it !

I don't think there's really any question about the DM drives, there's not much reason to doubt that they're 5900 rpm drives.

On a different note, the two slower drives are still on the first read phase of badblocks while the other three are on their second write phase. I'm wondering why they're performing so slowly :confused:
 
I got my drives today and they are indeed 5900 RPM. At least this is the value reported by the disk. I also did a frequency analysis using a contact microphone which showed a peak at 99 Hz (5940 min‾¹).



Now the remaining question is why Seagate told my distributor it was a 7200 RPM drive.

I'm not complaining though, 5900 RPM is what I wanted and it's nice that they were at least right about the 2 year warranty.
 
Thanks for the test. What softwares report is meaningless because it comes from a database.
 
well for you guys who will try to use them with HWRAID good luck and fill me in if its 7ms or changeable that wont reset on a powercycle.
 
Roughly 6% (10 hours) into the badblocks test I've got a Runtime_Bad_Block count of 1 on two of the disks. What does this mean? Is this a reason to be concerned or "normal"? I can't find any explanation for what runtime bad blocks means. Badblocks hasn't reported anything though.
 
I don't have any of these new 4TB but if you plan on using them for a NAS/storage go with the lower RPM if your concern about the noise of the enclosure. I have 7200 3TB seagates and man they are running hot to the touch in a RAID enclosure and would not be able to keep the NAS running 24/7 without it revving up the fans bothering light sleepers.
 
Best Buy has external 3TB for $120.

Keep in mind everything is relative. In 1975 10mb HDD cost $1m; I had a PC from 1992 with a 850mb.
Is a 16GB USB drive going for $9.95 priced low? Compared to what? A JAZ drive? I had a ZIP drive w/100mb discs costing $10 each.
For my first digital camera I bought a 128mb SD for $25.

BTW, I use a 500GB external HDD. with 300gb still free. I would be afraid to place "all my eggs in a single basket". I would get 2 2TB drives instead of 1 4TB.
 
Thanks for the test. What softwares report is meaningless because it comes from a database.
Only if the software is clueless.

The drive (firmware) itself will report its nominal rotation rate, in the result of a Identify Device command.

Also, one can measure/deduce the rotational speed, with user-level code, by doing read transfer timings and noting the track-to-track transition points.

--UhClem
 
Last edited:
Just grabbed a pair of these at Best Buy at $180 per... One is firmware CC43, the other is CC51. I can't find anything about CC51. I haven't taken either out of the enclosures yet so I'm not sure about models, but based on a few early read tests (random access 512 B and 64 KB) they seem to perform the same and are 5900 RPM, so both are likely DMs.

Update: Both are DMs. The CC43 is warrantied through April 2015, the CC51 through May 2015.

Update 2: I just took both drives out of the enclosure... It's strange that the newer CC51 DM had no warranty-voiding temper-proof tape binding it to the metal carriage yet the older CC43 DM does. I don't know what to make of that considering both drive SNs (not enclosure SNs) show a warranty of their own.

Update 3: The tape left no residue/mark so perhaps it was just to secure the drive.

More significantly perhaps, the CC51 drive says "Product of China" from site "WU" (Wuxi) and the CC43 drive says "Product of Thailand" and site "TK." That makes me a little uneasy about the reliability of the newer drive. The Chinese drive has a date code of 13296, the Thai drive 13213.

Does anyone else have a Chinese DM? I obviously haven't bought new HDDs in a while, not since the flood (and no Seagates since '00 or 01), so I didn't know to check the label for Thai vs Chinese :/ I'll exchange it tomorrow.
 
Last edited:
Best Buy has external 3TB for $120.

Keep in mind everything is relative. In 1975 10mb HDD cost $1m; I had a PC from 1992 with a 850mb.
Is a 16GB USB drive going for $9.95 priced low? Compared to what? A JAZ drive? I had a ZIP drive w/100mb discs costing $10 each.
For my first digital camera I bought a 128mb SD for $25.

BTW, I use a 500GB external HDD. with 300gb still free. I would be afraid to place "all my eggs in a single basket". I would get 2 2TB drives instead of 1 4TB.

850mb hard drives did not exist in the affordable Consumer PC sector until at least 1995...did you mean to say 85MB?
 
Maybe not. Look up the hot deals thread, some people have pulled them up and looked up the serial numbers and the drives have warranties on them. Actually the earlier ones came with 7200rpm drives and the warranty on the drive itself was longer than the warranty on the drive still in dock. But yeah I was at Costco ready to buy 4 or 8 but ended up deciding to just wait 6 months - year or whatever it takes for the internal 4GB drives to get down to 150.

And I've pulled a 3TB seagate from an external enclosure, typed in the serial number and it said "no warranty" oem drive. Just be careful when you pull apart the enclosure so no tabs are broken and stash it away. So if you do have to RMA, you can put it back and they won't know.
 
After using these drives for a while I have really started to take notice of the noise they are making. They are much louder than my WD Greens when seeking. Is there any way to silence them? It seems that AAM is not an option, but are there other options?
 
After using these drives for a while I have really started to take notice of the noise they are making. They are much louder than my WD Greens when seeking. Is there any way to silence them? It seems that AAM is not an option, but are there other options?

I haven't bought a Seagate in 10 years and now I remember why. These drives are hot and loud. I have DMs, can't imagine the 7200 RPM DXs.

These drives are louder and run hotter than my 320 and 640 GB 7200 WDs and, obviously, my Samsung and Hitachi .5, 1 and 2 TB 5900/5940s. It's too bad Hitachi is no longer an option. It's now a matter of choosing the least awful between Seagate and WD. Seems Seagate is the least awful at the moment.
 
These drives are hot and loud. I have DMs, can't imagine the 7200 RPM DXs.

What kind of temperatures are you seeing?

I saw a quote on NewEgg that stated:

Pros: Its got Seagate's latest platter technology , it runs cooler than WD Red's and its got the best price per GB for an internal drive yet.

I am looking to upgrade my home storage with eight drives, this was the drive I was looking at, but your comments have me worried.
 
Louder, yes for sure. Not that much hotter though. My 5 Seagates have an average temperature of 31°C while the WD Greens are at 28°C. The WD disks are sitting in a 4-in-3 bay though, and the Seagates are in 3-in-2 bays so the cooling is better for the WDs. Apart from the one WD which is sitting in a 3-in-2 and is the hottest of them all at 34°C. So I wouldn't say that the ST4000DM000 runs particularly hot. I can imagine that a 5 platter 7200 RPM drive does though.
 
I wouldn't say that the ST4000DM000 runs particularly hot. I can imagine that a 5 platter 7200 RPM drive does though.

While I'd rather it be in the 25C - 30C range, I can live with 30C - 35C.

I saw this is another NE review:

The DM is 5900 RPM, 64 MB cache and 4x1000GB platters.
The DX is 7200 RPM, 128MB cache and 5x800GB platters.

Writes were about the same on both drives. Reads are a little faster on the DX. The DX runs MUCH hotter, as much as 38% in one of my tests (53C vs 38C under heavy load with fans on low).

53C vs 38C... wow.
 
But no warranty since you violated the terms by ripping it out of the casing. If you pull the drive and it dies a day later , you're fucked.

According to the thread over @ SD, people contacted Seagate and they said the drives that had serial numbers that showed up on their website as having a individual warranty, WERE covered even though they were removed from the case. I know it seems weird
 
Back
Top