Samsung F1 series hard drives w/1TB model

Jose Hidalgo · Post by **Jose Hidalgo** » Sat Feb 02, 2008 6:08 pm

Luminair wrote:
Jose Hidalgo wrote:As requested by lm, here's the way we have used to test our drives with Debian
Debian (GNU/Linux) isn't Unix, and a read/write test does not check all areas of the disk. The master cylinder, for instance, is not checked by that test.

Yes, but that was not the purpose of our test. I was simply answering someone's request, not trying to justify anything.

The purpose of our test, in our particular case, was to check the R/W operation of a complete chain with on-the-fly encryption/decryption (xfs, dm-crypt, mdadm, and the chosen kernel). It's a pretty good intensive R/W test anyway, so I consider it useful.

But if you simply want to check a whole disk you have the badblocks command : http://linux.die.net/man/8/badblocks

Fedor · Post by **Fedor** » Sat Feb 02, 2008 9:21 pm

By the way I have bought 4 750GB versions for a RAID5 array and they all came back with the MC error (reading previous posts it seems there is a lot of confusion what this MC stands for, from media check to master cylinder, but afaik its maintenance cylinder). They had no other errors, and I did run the full surface scan. Performance was on spec, so then I proceeded to create the RAID5 array, and after I dumped my existing 5 x 500GB RAID5 data on it. I .sfv checked the data and things are still checking out but so far so good.

So it's not just the 1TB models that have this issue, and I figured I'd report my experience thus far to add another account of the drives seeming fine in every conceivable way except for that stupid MC test.

metro_88 · Post by **metro_88** » Mon Feb 04, 2008 8:33 am

Thanks Fedor for letting us know the 750's (I'm assuming the HD753LJ) are having similar problems.

I've been running two HD103UJs all weekend on RAID 1 without any problems. These were both drives that failed the "Check M.C" test.

I would advise anyone who's drives fail any tests other than "Check M.C" to RMA for a new drive. For instance, one of my first two drives failed the surface scan as well.

Sascha · Post by **Sascha** » Mon Feb 04, 2008 8:43 am

Just wanted to add my experience here as well:

I recently ordered two HD753LJ (750GB) drives for use in a small RAID1. After experiencing some weird problems (mainly copy operations freezing up the machine) I ran Samsung's HUTIL and that also gave me the ever so popular "M.C. Error" on both(!) drives as well as Surface Scan errors on one of the drives. I'm currently doing an ERASE using HUTIL and will run the surface scan again after this. I'll post my results here once this is done.

Cheers,
Sascha

metro_88 · Post by **metro_88** » Mon Feb 04, 2008 8:53 am

Thanks Sascha.

The Hutil erase operation did not affect my surface scan error on my first HD103UJ. Let us know if you have different results.

andyb · Post by **andyb** » Mon Feb 04, 2008 11:09 am

"metro_88"

My drive failed the surface scan in quite a big way, I have (twice now) installed Windows XP, Drivers and software on that area of the drive with no issues at all, I tested the drive on 4 chipsets all of the info can be found in my previous posts.

I am currently trying to get some joy out of Samsung, on hold with the HDD team again, and yet again have been promised a call back.

Andy

Luminair · Post by **Luminair** » Mon Feb 04, 2008 11:48 am

MC error debate aside, I think you owe it to yourself to not settle on a drive with surface errors.

metro_88 · Post by **metro_88** » Mon Feb 04, 2008 12:19 pm

Andyb, thanks for your continued persistence with Samsung. Let us know if you get any contact information and I think there'll be more than a few of us that will be contacting Samsung.

I'm glad you've not had troubles with your drives Andyb. My comments about surface errors were to let people know that ALL of these drives have Check MC errors, but not all have surface scan errors. I, personally, would RMA, but who knows, Hutil may be misreporting these as well.

andyb · Post by **andyb** » Mon Feb 04, 2008 1:57 pm

MC error debate aside, I think you owe it to yourself to not settle on a drive with surface errors.

I'm glad you've not had troubles with your drives Andyb. My comments about surface errors were to let people know that ALL of these drives have Check MC errors, but not all have surface scan errors. I, personally, would RMA, but who knows, Hutil may be misreporting these as well.

My point was not specifically to do with the Smart MC error, but to do with the fact that there are 0 relocated sectors, the drive is in perfect health according to SMART data, and I can use the drive fine even though it is supposedly FUBAR according to HUTIL.

If the drive was as phuqt as HUTIL says it is then I shouldnt have even been able to complete an install of XP, let alone copy large amounts of data from the drive and back to it without issues.

WTF are "Sector ID not found" errors. according to a few links from Google that is an unrecoverable error that should mean mass data loss. 5% of the first 15% of my drive gets this error, in no way should this be usable, but it is, this can only mean one thing the HUTIL tool is faulty.

Andy

andyb · Post by **andyb** » Tue Feb 05, 2008 8:20 am

Update from Samsung

They have admitted that their software (HUTIL version 210) DOES NOT work with the 1TB F1. The gentleman was not at liberty to tell me whether their software does not work with the entire F1 line or not.

Samsung are working on a new utility, a new version of HUTIL, or a piece of software specifically to test the F1 line (or just the 1TB model).

Keep your eyes on the following page for more information or the new utility.

http://www.samsung.com/global/business/ ... HUTIL.html

Andy

Jose Hidalgo · Post by **Jose Hidalgo** » Tue Feb 05, 2008 9:34 am

Thanks andy.

So that's exactly what I suspected : the problem is coming from the software, not necessarily from the drives.

So people, stop complaining and go buy a bunch of F1's : they are worth every penny of their price.

nevion · Post by **nevion** » Tue Feb 05, 2008 4:23 pm

I'm glade that's so - wasn't looking forward to another mass migration of data if it turned out they lending themselves to being faulty.

Mass production f*ckups do happen though and electronics (especially those with moving components) have many points of failure in manufacturing processes - something to keep in mind in a situation like this or rather any new hardware acquisition that hasn't proven itself yet. These drives are the first generation models after all.

NiToNi · Post by **NiToNi** » Wed Feb 06, 2008 3:37 am

Before everyone blames HUTIL v2.10, I thought I'd add my own experience with the F1 1Tb drives.

I bought five of them (5x 1Tb) here in the UK back in the middle of January, which all had serial numbers within a range of fifteen-twenty numbers of each other.

All five showed an ECC error on the M.C. test and four had also errors on the Surface Scan test. I low level formatted one of the drives that had both errors but it did not make a difference, so I couldn't really be bothered to go through formatting the other four.

I RMA'd within the week and took delivery of five new drives with totally different serial numbers, although still within a close range from each other (so I assume all came off the same batch).

Predicatbly, they all failed the M.C. test at which point in the I stopped the test, i.e. before the thorough Surface Scan as I knew I was not going to keep faulty drives (and honestly because I was out of patience with Samsung and the whole thing).

Full refund from retailer so all well there. Still put off though by having had to go through the very time-consuming hassle of funding, taking delivery of, installing, testing, communicating with tech support, repacking, resending, having to wait for refund for five faulty drives... twice... for nothing

Most interestingly however was that this was far from a academic issue for me; I could not get the harddrives to work with my Areca ARC-1231ML PCI-E raid controller card, not even when connected one by one, and the card configured for JBOD operation. The card recognised the drives alright but always hung on boot after the five-second countdown as part of its BIOS boot sequence and just before the O/S was supposed to load (If the drives were switched on in Windows Vista, they were visible in Disk Manager as long as they were connected straight to a SATA port on my motherboard).

I have no such problem with my two Barracuda 7200.10 750Gb drives, which makes me believe that these F1 HDDs suffer from more serious problems other than not being compatible with HUTIL v2.10.

I hope Samsung will sort this out and I am holding off buying other drives, at least till another manufacturer introduces a 3x333Gb platter product.

Cryoburner · Post by **Cryoburner** » Wed Feb 06, 2008 5:03 am

I just thought I'd point out that Newegg has their F1s on sale right now. The 750GB is available for $149.99 and the 1TB for $269.99, both with free shipping.

Still no sign of the 320GB or 640GB units though. : \

Fedor · Post by **Fedor** » Wed Feb 06, 2008 5:04 am

Well, not all RAID cards work with all HDDs well by a LONG shot, so it's very possible they were simply incompatible. For example, even WD's Raid Editions had trouble on Adaptec cards in the past - at least with the Raid Editions they didn't let it slide and released a firmware fix.

So why am I defending the F1s when my four 750s showed the M.C. error as well? Well, they are working fine in my PERC5i (LSI 8408E) in RAID5, so that's definitive proof that we have no proof of whether they have as you said "more serious problems"

As for other 3x333 drives, the current WDs coming out with 320GB platters should fit your needs quite well

htgamer+ · Post by **htgamer+** » Wed Feb 06, 2008 1:31 pm

I noticed that Samsung has enterprise version of the F1 drives. The 750gb version is called the HE753LJ.

Aside from the ususal PR & marketing about more robust etc., is there really any difference bvetween this model and the desktop models. Is the "E" version just a cherry picked (or same as) "D" version. Does anyone know if the "E" versions have been afflicted with the same problems as the "D" version. I am new to Samsung drives and from what I have read above this is pretty scary compared to the good luck I have had with those noisy Seagates.

metro_88 · Post by **metro_88** » Wed Feb 06, 2008 1:47 pm

After some simple Googling, I see Samsung has a 1TB RAID class drive as well, the HE103UJ.

Link to it on Samsung's site.
http://www.samsung.com/global/business/ ... &ppmi=1162

I can only find it listed in a few places and it seems to be about $100 USD more than the HD103's.
http://www.google.com/products?q=HE103U ... ucts&hl=en

I haven't the slightest idea, however, what Samsung (or any other manufacturer) does to these drives to make them RAID/enterprise class. Maybe more sophisticated onboard electronics?

I'm sure someone more knowledgeable than me can answer.
[/url]

Jose Hidalgo · Post by **Jose Hidalgo** » Wed Feb 06, 2008 2:16 pm

It's quite simple : they do reduce the drive's recovery time, so that the drive can fit in a RAID array.

For normal drives, when there's a R/W error, the drive can enter a "deep recovery cycle" that can take up to 2 min. During that time, the drive becomes unresponsive.

Most RAID arrays do not tolerate a drive being unresponsive for more than 8-15 seconds. If it does happen, the drive is considered "faulty" and ejected from the RAID array. Plain and simple. Of course there are workarounds, but that's not the subject here.

So from that point of view, a RAID-edition drive is simply a normal drive with a recovery time that has been truncated by firmware operation to less than 8 seconds (they dare call that "Time Limited Error Recovery", like if it was a new feature

). It's as simple as that.

Oh, and of course, there's the MTBF that goes up from 800.000h to 1.200.000h or so, with no precise explanation. Like if something secret had been miraculously added to the drives.

So well, if you believe that 400.000 more hours are worth $100, then go for them. But do not expect anything more.

Oh BTW, here's an explanation from WD. It's the same thing I've just said, put in more elegant words

:

Q: Regular 7200 RPM desktop drives run fine in RAID environments; why do I need these drives?
A: Unlike regular desktop drives, WD RE SATA and EIDE hard drives are engineered and manufactured to enterprise-class standards and include features such as time-limited error recovery that make them an ideal solution for RAID.

Q: What is time-limited error recovery and why do I need it?
A: Desktop drives are designed to protect and recover data, at times pausing for as much as a few minutes to make sure that data is recovered. Inside a RAID system, where the RAID controller handles error recovery, the drive needn't pause for extended periods to recover data. In fact, heroic error recovery attempts can cause a RAID system to drop a drive out of the array. WD RE2 is engineered to prevent hard drive error recovery fallout by limiting the drive's error recovery time. With error recovery factory set to seven seconds, the drive has time to attempt a recovery, allow the RAID controller to log the error, and still stay online.

lm · Post by lm » Wed Feb 06, 2008 3:06 pm

Jose Hidalgo wrote:Oh, and of course, there's the MTBF that goes up from 800.000h to 1.200.000h or so, with no precise explanation. Like if something secret had been miraculously added to the drives.

Failures during the first couple of months of the drive are more common.

With some additional stress testing, they are able to weed out some drives that would die early. That improves the MTBF numbers.

I think you can achieve the same effect by not keeping any mission critical data on the drives for the first 3 months, or just make sure everything's well backed up. But commercial users probably want to pay for someone else to deal with it.

Fedor · Post by **Fedor** » Wed Feb 06, 2008 6:05 pm

And other than TLER, the only other definite tangible benefit is that other than Seagate, which provides 5 year warranty on all their drives, the Raid Editions typically come with 5 years compared to the manufacturer's usual 3.

dhbn · Post by **dhbn** » Wed Feb 06, 2008 6:34 pm

Gentlemen:

I just bought a 750 GB Samsung HDD (HD753LJ). After I installed it (second drive in a HP Pavilion a828.dk desktop PC), about every second time I start my computer the Samsung drive has disappeared (it is partitioned and formatted, of course). When this happens, it is absent in Windows Explorer as well as in Computer Administration (or whatever that utility is called in English - the one where you can partition drives, change drive letters etc.). Next time I reboot, the drive re-appears. Next time, gone - and so on. When it is present in Windows, it seems to work fine.

This problem was the reason why I ran Hutil. I got the same error that everyone seems to get - the ECC error in connection with the Check MC test - and I thought that was the cause for the problem. I found this thread when searching for information about the error. Now I see that Hutil is apparently falsely reporting the MC ECC error for the 1 TG HDD, perhaps the same is the case with the 750 GB drive. In other words, I don't know if the problem with my HDD disappearing in Windows has any connection to the MC error.

I don't know if this has any relevance, but the first time I booted my computer after I installed the Samsung drive, the receiver for my wireless Logitech mouse stopped working. It took a couple of reboots to get it to work again. This hasn't happened since - now it's only the HDD that disappears half the time. Oh, and I tried disconnecting the receiver and rebooting to see if the drive would stop disappearing, but it still did. I also tried another SATA cable; that made no difference.

Another thing I noticed is that when I first ran Hutil, it didn't find the Samsung drive. I then disconnected my primary drive and moved the Samsung SATA cable to the primary drive's SATA port on the motherboard, after which Hutil could find the drive. Does this indicate an error in the SATA port on the motherboard? Does it make any difference which drive is connected to which port - I mean, will the computer always try to boot from the HDD that is connected to the primary port? If not, I could try switching the HDDs, but I'm afraid to try if it might mess up the Windows installation somehow.

Can anyone help me with what is going on? Perhaps I am the one unlucky guy who in fact has the MC error that Hutil is reporting for everybody! Any help appreciated.

Michael Sandstrom · Post by **Michael Sandstrom** » Wed Feb 06, 2008 8:18 pm

Switching the HDs should not cause any problem. In the boot sector of the BIOS you should be able to choose which drive to boot from. I don't have experience with your HD but my two Samsung HD501LJ units are slow to establish communication and I have to Ctrl Alt Del every time after a cold boot with my VIA SATA1 controller.

Luminair · Post by **Luminair** » Thu Feb 07, 2008 4:28 pm

dhbn wrote:Another thing I noticed is that when I first ran Hutil, it didn't find the Samsung drive.

It sounds like you have a bad cable or a bad motherboard, so you should take your problem to a different thread

dhbn · Post by **dhbn** » Fri Feb 08, 2008 1:44 am

Luminair, I'm really not sure about that! I tried switching both cables and SATA ports on my motherboard so that the Samsung HD got the cable and port that my boot HD used to have and vice versa, and the PC continued to boot just fine, so it had no problems "seeing" the boot HD. On the other hand, the Samsung kept disappearing...

I've now also tried the Disk Manager program that Samsung have on their website. It should be able to solve problems like this one, but hasn't. I'm mystified!

Anyway, I was just hoping that someone might have heard about this before and know how to fix it... I'll probably return the HD and so won't need to bug you guys anymore in this thread!

Kaleid · Post by **Kaleid** » Fri Feb 08, 2008 1:58 am

Another thing I noticed is that when I first ran Hutil, it didn't find the Samsung drive.

I had the same problem. 2.10 does not find my T166 cr-100-12 drive so I had to redownload 2.03

My other t166 has been sent to be repaired/exchanged since about 1600KB clusters were damaged.

Couldn't even run hutil's erase.

andyb · Post by **andyb** » Fri Feb 08, 2008 7:53 am

Another update from Samsung.

The 750GB is also affected by the "bugs" in the HUTIL 210 software.

The gentleman who phoned me said that the software had bugs in it, and should not be used for the 750GB and 1000GB F1 drives, and likely any others in the F1 range (if anyone has any of them).

He suggests that people assume that there is not a problem with their drives unless they have problems or suspect a fault, and test the dirves using a different utility if they want to do any testing such as "HD Tune" (my suggestion).

Version 211 of the software is being made, but due to the main Korean holiday of the year this will likely take 2-3 weeks, keep your eyes on their website.

I would personally suggest that no-one uses the v210 software at all as I have had many issues with other drives as well, such as not evenb being able to scroll down within the software.

Andy

DuncanG · Post by **DuncanG** » Fri Feb 08, 2008 1:56 pm

Michael Sandstrom wrote:I don't have experience with your HD but my two Samsung HD501LJ units are slow to establish communication and I have to Ctrl Alt Del every time after a cold boot with my VIA SATA1 controller.

I have the same problem with my Gigabyte GA-MA69GM-S2H motherboard and HD103UJ. The BIOS never recognises the disc on a cold boot, only on subsequent boots. Can't get any help from Samsung.

Keto · Post by **Keto** » Fri Feb 08, 2008 11:15 pm

andyb wrote:Another update from Samsung.

Version 211 of the software is being made, but due to the main Korean holiday of the year this will likely take 2-3 weeks, keep your eyes on their website.

Andy

Ask them to make it work with SATA optical drives!

Fedor · Post by **Fedor** » Sun Feb 10, 2008 1:14 am

I was 99% sure the program was misdiagnosing my 750s that I mentioned earlier, but it still adds to my peace of mind knowing they've admitted it.

My bet is that those with surface scan errors really did have bad drives though, and it seems thats mostly the 1TB models, perhaps due to their thicker platters.

andyb · Post by **andyb** » Sun Feb 10, 2008 6:11 am

My bet is that those with surface scan errors really did have bad drives though

According to Samsung, you just shouldnt believe whatever Hutil 210 says about those drives. I am using my drive and have been for several days with no issues, by drive was smothered with errors.

Andy

Samsung F1 series hard drives w/1TB model

My $0.002