Hard Drive problem on X58A-UD3R

Toggle sidebar Toggle sidebar
D

discodave

Distinguished
Jul 24, 2010
13
0
18,510
Hi

I've recently completed a new build based on a X58A-UD3R thanks to this forum and it's very helpful members.

I have a problem though with one of the hard drives. Both are brand new.

I have 1 SSD and 2 Samsung HD103SJ Hard Drives connected to the ICH10R SATA controller set to AHCI mode. The Samsungs are just connected as single drives not RAID.

The first Hard Drive seems to be working fine but I have had problems with the second twice now. The first time I had them installed in RAID 0 but the Intel monitoring software flagged up a warning on one of the drives. I took them out of RAID and reformatted as separate drives. Again though while copying files to the drive it went down and Windows lost it. It came back after a couple of reboots but obviously all is not well.

Firstly could my set up be causing any problems?

Second, if not, what is the best process to find out if the drive if faulty or something else.

I tried the Samsung Hutil utility but it does not recognise this model, maybe because it's too new.

I need to RMA is soon if it's faulty.

Any suggestions would be much appreciated.

Many thanks

Dave

Other System details.
X58A-UD3R
i7 920 OC @ 4Ghz
6Gb Corsair RAM
Bundle from Scan.co.uk
OCZ vertex 2 60Gb SSD
450w Corsair PSU

 
Hi DiscoDave,

A fairly easy way to try to pin down whether this problem is with your Samsung HD103SJ Hard Drive or the GB motherboard, is to take the HDD, and just connect it to a non primary SATA port on another computer, format it, and copy some files to it, and just let it spin for several days.

You could turn on the S.M.A.R.T reporting BIOS function, that might report a problem, like the spin up time, or seek reporting, or temperature.
There is also a free S.M.A.R.T applet available on the web that can do that also called DiskCheckUp which you can download, and uninstall when not needed anymore.

Also if you have a recent version of CheckIt Diagnostics from Smith Micro (sometimes included on a Symantec CD) you could check the hard drive/s read/write ability with it.

If you have trouble with it on a separate desktop, or if that BIOS fails to see the drive, you know it is the HDD, giving you the migraine.

Hopefully that will help you determine which critter to RMA.

John VK
 
Hi

Thanks for all of this. I will try all over the weekend.

To add to this in response to some of these comments.

Last night I downloaded the trail of HD Tune Pro and did a surface scan of the drive in question and it discovered a couple of bad blocks. There were some warnings flagged in the health tab as well but I can't remember right now what it said which isn't much use I know but I'm not with the machine right now. Interestingly and annoyingly it had a health warning for the SSD as well which I though was operating fine.

The SSD and the other Samsung drive to my untrained eye seem to be working fine and I would not have thought otherwise had I seen the warnings in HD tune.

I've already put the drive in another computer to try and test it with the Samsung Hutil and it was recognised by the bios fine. just that Hutil didn't recogise this model.

Could i use the HD Tune trial to test it effectively seeing as it's already installed. Obviously not much hassle do download the other software as well. I will try ATTA Disk benchmark as well.

A couple of questions.

1. Could I have done any harm when initially putting the 2 HDDS in raid 0 and only quick formattting? Would anything fix that? Reformatting with HD tune and zero all or something.

2. doesn that fact that the SSD has a warning from HD tune point to another issue?


I will first test the faulty drive fully on another machine.

What is the ideal way for me to reformat the drive? Is a windows full (not quick) format sufficient or should I something else?

Once again thaks so much for the help.

 


Thanks bilbat

SSD is on ICH but not with these drivers.

I installed the drivers from the Gigabyte CD and made sure they were all up to date is all, just the drivers section not any of the additional gigabyte software. I'm seeing quite a lot about Gigabyte software not being too great.

Is there anything else I should uninstall and replace with manufacturers versions?

I'm a little confused about the need for a floppy? I installed Win7 to the SSD on the ICH set to AHCI without the need to f6 driver from a floppy. There is a floppy driver diskette for Win 7 here though.

Maybe I just need the allOS version in this instance?

 


[1] Harm ~ no ; fix ~ running Check Disk w/both options will recover/flag as bad so no data is written to the "bad sector" ; full reformatting does the 'same' as Check Disk.

[2] All of your new SSD + HDD having problems is a huge red-flag of the on-board disk controller having a problem.

Bottom-line, if you can use the SSD + HDD on another computer then it kinda R/O the disks to me. I "guess" the reason that I feel the way "I do" is because I have been scouring the forums for {X58A-UD3R} with RAID issues, I wasted too much of my time and second guessed myself to the point of doubt "and" I hate to see someone experience the frustrations I did...

Also, at the end and before I RMA'ed the last of the 10 PCs I still had problems with the HDD as non-RAID, I even took a Delta fan and aimed it inches away from the MOBO controller ~ no joy.

The ATTO in my case forced (stressed) the MOBO to turn-off from what Gigabyte "assumed" to be overheating protection kicking in, and in my case the problem became progressive {worst over time}.
 


~ I went through ALL of this "stuff" (x10) and in simple terms "plugging-in drives, installing OS, installing drivers, w/default BIOS" ~ will make your PC work fine. Tweaking to AHCI means zip (0) to making it "work".

To save you some time, meanwhile create a ticket with Gigabyte - http://gigabyte.com/support-downloads/technical-support.aspx
Continue the forums, tweaking, reconfiguring, installing drivers, formatting, reformatting, Clear CMOS, use all possible BIOS settings, (rebuilding the array 15+ times), etc...you get the idea... :pt1cable:

One thing that I did NOT try and if you're game, place the 2 HDD on the GSATA3 ports ; not as efficient but it'll be a different controller. In my world "I trust" only Intel RAID so even if the GSATA3 worked it was unacceptable "to me."
 
Seven supports AHCI with 'native' drivers, meaning the drivers are already in both the OS and the installer kernel.

The 'allOS' file is all you'll need to run - it will 'freshen' your drivers, and install the Intel Rapid Storage Technology manager. Try to ignore jaquith - in addition to being semi-literate, unreadable, and very confusing, he consistently misses citing the fact that he received a shipment of motherboards in crushed boxes, and tries to somehow implicate GB in his failure to refuse shipment. In five years of using them, I have never seen a failed ICH - and their max temps are all in the 100-105°C range
 
Thanks again for all of your comments. I'll put it all to the test.

I'll post later what HD tune is saying about the SSD and if anyone has the time to comment I'd be very interested to see what the problem might indicate and if there is a real problem with the SSD.

thanks

 
The "more" stress test on the SSD {read/write} the better ; BSOD or a sudden shut-down indicates a "serious" issue. I had no issues with the {primary} SSD, the problems were with my HDD, the ATTO always was fine on the SSD, but running ATTO on the RAID + as non-RAID caused the system to suddenly shut-down and/or BSOD.

Corruption ~ on new equipment, in your case, indicates a bad on-board controller especially if the corrupted sectors are recoverable. The ultimate way to check is reformatting the HDD and comparing the (2) HDD info. Corruption of ALL of your "drives" is 0% making the MOBO the cause!
 
Hi Again

Once again thanks for all your comments, I've tried a few things.

Summary:

One of my Samsung HDDs still has warnings from HD Tune even when connected to a different PC. The Other is working fine on the new GB Mobo. The SSD is probably fine on the new GB MoBo but I need to test to be sure.

More detail:

I put the possible faulty drive in another PC. I reformatted and ran chkdsk with both boxes ticked on the drive in Win XP (by right clicking drive/Properties/tools/Disk Checking).

It ran through chkdsk OK and did not flag up an errors, it just said checking finished or similar, not sure if it fixed anything as it didn't say. So to be honest I'm not sure what that means?

But, I ran HD tune on the same alternative PC after chkdsk had finished and it's still flagging up the same warnings as it did when connected to my new BG MoBo. Which would say to me it's not the GB MoBo.


This is as screen grab of HD tune running on the alternative PC. These results are the same as when it's connected to my new GB MoBo.


My other Samsung Drive does NOT have any errors at all, HD tune is saying all is well when connected to the new GB MoBo.


This is a grab of HD Tune running on the new GB Mobo on the good HDD.


Now HD Tune is is flagging a health warning on my SSD connected to the new GB MoBo but from looking at a few posts HD Tune is not intended for SSDs so this is not a good way to judge it. I saw a response from OCZ support to the same warning in HD Tune and they said that HD tune should not be used to test SSDs. Saw a similar one for an Intel SSD where HD tune flagged warnings but Intel's own utility said all was OK. I can't find a utility form OCZ though.

This is what HD Tune is showing for the SSD anyway.


This is a grab of HD tune on the new GB Mobo on the SSD.


Questions:

Do these HD Tune health readings for the Samsung HHD point to it being faulty?
Is there anything I could do to fix these or should I RMA now?

How should I best test the OCZ SSD to be sure it's OK when OCZ don't seem to have a testing utility?

Hopefully this is pointing to the Mobo being OK.

Any comments much appreciated.

Cheers


 
OK Just tried copying a large folder of MP3s to the faulty drive on the alternative PC and almost immediately win Xp lost the drive again.

Surely this must be faulty?

Only thing I haven't tried yet is a different cable. But surely formatting and chkdsk wouldn't have worked or there would be other problems if it is a faulty cable?

I'm surprised that chkdsk didn't have problems with the drive though (see above post)

I'm still a little confused.

thanks
 
Not all that complex, really - hopefully, I can explain it in a non-complex way [:fixitbil:9] !

The reallocated sectors and calibration retry counts are likely related - and both of them bad news. The drive works by pivoting a little head in and out over tracks that are like the grooves in a record (eek - just thought - there's a couple generations out there who've never seen a record [:bilbat:6] ); then each groove is divided into portions of data called sectors. Obviousy, the heads have to be able to 'find' the grooves and sector markers reliably - as there's no physical groove to keep it 'captured' - that's all done electronically with feedback loops. The drive calibrates itself to be able to 'find' these things - and if the calibration fails - it's in trouble! If the cal fails for a specific spot, it can't reliably 'find' that spot; and if, for any reason, it gets a command to 'go here', and it either can't find 'here', or there's nothing intelligible 'here', it re-maps 'here' to 'there', and remembers it - that's a 'reallocated' sector!

Now, as incredibly dense as modern recording media/technique is, it's not unacceptable to have a few flaws in the process - it's like the occasional bad pixel in large monitors - one or two 'passes' QC. I've seen a lot of 'perfect' drives; but I'm never concerned with a sector or two that's had to be remapped. When the number gets large, however, or even starts 'growing', it's indicative of impending failure - especially when accompanied by your calibration errors... The reason format or checkdisk aren't 'seeing' these problems is that the drive electronics are built to 'hide' this - check disk says 'check here', but the drive has reallocated 'here' to 'there', so checkdisk 'sees' 'there', and 'there' is fine - so checkdisk never 'throws' an error.

I think the reason behind not using conventional tools to analyze an SSD is that they use this 'remap' ability as part of their standard mode of operation - they have to, as the number of times an MLC memory cell can be rewritten is limited. The drive never 'wants to' re-write a cell for the fourth time, if there are cells available that have only been written twice...
 
Hi DiscoDave,

When you have a new HDD that has warnings like this Samsung has, send it back for a replacement, or possibly the retailer will replace it. It will only get worse. I agree will Bill, don't think about it anymore, it is a problem and not a risk worth taking.

I don't know if there are any quirks about checking the SSD for errors. The SSD's move sectors around so as to use the drive more evenly, and maybe checking them is a little different.

You are using the trial version of HDTune Pro, which is good. The HDTune website says it does check SSD's.

You might consider waiting till Monday, and call OCZ tech support and ask them about the best way to check SSD's & what that one warning means. Their tele # is (800) 459-1816, or in San Jose (408) 733-8400.

One thing I usually do when a shareware or trialware utility has really helped me, is to consider buying the application for my toolkit for future use, and as appreciation for it's help.

Will be anxious to hear what OCZ has to say about SSD "health checkups"

 
Answer Summary ~ There is a difference between "Write Errors" and "Physical Errors" having (2) new drives with Sectors missing from found "Write Errors" can be from a few sources (e.g. MOBO Controller or HDD/SSD Controller). As I mentioned a comparison from before and after a Full Formatting or Disk Repair is the only way to R/O "Physical Errors".

"IF" you didn't have a X58A-UD3R MOBO I still might be swaying towards the MOBO, but since you do have the X58A-UD3R AND there have been numerous posts similar I still feel it is your MOBO as the root cause.

RE: "alternative PC and almost immediately win Xp lost the drive again" clearly would indicate a bad HDD unless the Journal file was ALREADY corrupted. Out of curiosity alone, I ran HD Tune on the oldest Test XPSP3 I had running almost non-stop for 5 years with a WD 300GB SATA1 HDD, ran the full ~1 hr test ~no errors and health OK.

footnote: EVERY-TIME I install a new SSD/HDD I perform a FULL Format. I do so to find and eliminate (minimize) problems, plus I routinely run full Disk Checking to maintain integrity of the data.
 
After a bit more forum searching I've found that the health reporting from HD Tune and the Reallocated Sector Count warning should be ignored for my OCZ Vertex 2 SSD. Something to do with the S.M.A.R.T info being incorrectly read. Seems this is very common complaint on the OCZ forum. Hopefully they will be releasing a new firmware update that will solve this sometime.

So my SSD is fine.

First HDD is fine and healthy.

Second HDD is faulty in two different PCs.

Motherboard is OK.

Happy days almost, except for having to return a drive but that's a damn site better than it being the Mobo. Just hope the replacement one is OK.


Janquith. I will keep an eye on it though and make sure things don't start degrading.

Bilbat. Latest Intel drivers installed and Rapid storage reporting all is OK and thanks for the explanations.

John. Hopefully some of the info above will answer your queries and yes I will purchase HD tune, it's been a real help.

Thanks again to you all for helping sort this one.

Dave


 
Always
welcomeho.gif
!
 
You must log in or register to reply here.

TRENDING THREADS

Latest posts

Moderators online

Share this page

COMPANY
RESOURCES
FOLLOW
Top Bottom