RAID 1 Fails when copying large amounts

[marc0523]

This problem has been around in my system since I purchased it but i just haven't bothered to get it fixed.

I have two Samsung HD103UJ 1TB as a RAID 1, this works perfectly 99% of the time.

When I copy or move large amounts of data around the raid will always fail though! this will happen apparently randomly with, i think, both drives failing (I have only just started to keep a log of when the drives fail).

i can, through the Intel Matrix Storage Console, mark the drive as normal, this will then start the rebuild, once, this is done it is as if the drive never failed!

I think this is a timing issue with the Samsung drive, I remember when i was working for Dell that we had loads of issues with Samsung drives and had to change an awful lot of them, many people i know wont go near the drives since.

Can anyone help me stop this happening in the future?


Thanks,
Marc.








Below is the Matrix Storage Controller Report:

System Information

Kit Installed: 8.6.0.1007
Kit Install History: 8.6.0.1007, Uninstall
Shell Version: 8.6.0.1007

OS Name: Microsoft® Windows Vista™ Ultimate
OS Version: 6.0.6001 Service Pack 1 Build 6001
System Name: SCAN-NINJA
System Manufacturer: ASUSTeK Computer INC.
System Model: P6T DELUXE
Processor: Intel(R) Core(TM) i7 CPU 920 @ 2.67GHz
BIOS Version/Date: American Megatrends Inc. 1003 , 12/05/2008

Language: ENU



Intel(R) Matrix Storage Manager

Intel RAID Controller: Intel(R) ICH8R/ICH9R/ICH10R SATA RAID Controller
Number of Serial ATA ports: 6

RAID Option ROM Version: 8.0.0.1038
Driver Version: 8.6.0.1007
RAID Plug-In Version: 8.6.0.1007
Language Resource Version of the RAID Plug-In: 8.6.0.1007
Create Volume Wizard Version: 8.6.0.1007
Language Resource Version of the Create Volume Wizard: 8.6.0.1007
Create Volume from Existing Hard Drive Wizard Version: 8.6.0.1007
Language Resource Version of the Create Volume from Existing Hard Drive Wizard: 8.6.0.1007
Modify Volume Wizard Version: 8.6.0.1007
Language Resource Version of the Modify Volume Wizard: 8.6.0.1007
Delete Volume Wizard Version: 8.6.0.1007
Language Resource Version of the Delete Volume Wizard: 8.6.0.1007
ISDI Library Version: 8.6.0.1007
Event Monitor User Notification Tool Version: 8.6.0.1007
Language Resource Version of the Event Monitor User Notification Tool: 8.6.0.1007
Event Monitor Version: 8.6.0.1007

Array_0000
Status: No active migrations
Hard Drive Data Cache Enabled: Yes
Size: 1863 GB
Free Space: 0 GB
Number of Hard Drives: 2
Hard Drive Member 1: SAMSUNG HD103UJ
Hard Drive Member 2: SAMSUNG HD103UJ
Number of Volumes: 1
Volume Member 1: Volume 1 - OS

Volume 1 - OS
Status: Degraded
System Volume: Yes
Volume Write-Back Cache Enabled: Yes
RAID Level: RAID 1 (mirroring)
Size: 931.5 GB
Physical Sector Size: 512 Bytes
Logical Sector Size: 512 Bytes
Number of Hard Drives: 2
Hard Drive Member 1: SAMSUNG HD103UJ
Hard Drive Member 2: SAMSUNG HD103UJ
Parent Array: Array_0000

Hard Drive 0
Usage: Array member
Status: Failed
Device Port: 0
Device Port Location: Internal
Current Serial ATA Transfer Mode: Generation 2
Model: SAMSUNG HD103UJ
Serial Number: S13PJ90QB07421
Firmware: 1AA01113
Native Command Queuing Support: Yes
Hard Drive Data Cache Enabled: Yes
Size: 931.5 GB
Physical Sector Size: 512 Bytes
Logical Sector Size: 512 Bytes
Number of Volumes: 1
Volume Member 1: Volume 1 - OS
Parent Array: Array_0000

Hard Drive 1
Usage: Array member
Status: Normal
Device Port: 1
Device Port Location: Internal
Current Serial ATA Transfer Mode: Generation 2
Model: SAMSUNG HD103UJ
Serial Number: S13PJ90QB15564
Firmware: 1AA01113
Native Command Queuing Support: Yes
Hard Drive Data Cache Enabled: Yes
Size: 931.5 GB
Physical Sector Size: 512 Bytes
Logical Sector Size: 512 Bytes
Number of Volumes: 1
Volume Member 1: Volume 1 - OS
Parent Array: Array_0000

Unused Port 0
Device Port: 3
Device Port Location: Internal

Unused Port 1
Device Port: 4
Device Port Location: Internal

Unused Port 2
Device Port: 5
Device Port Location: Internal

CD/DVD Drive 0
Device Port: 2
Device Port Location: Internal
Current Serial ATA Transfer Mode: Generation 1
Model: PIONEER DVD-RW DVR-216D
Serial Number: Data not reported
Firmware: 1.06

[Blitzen]

Its very rare for hard drives to have intermittent problems.
They usually fail or they dont..
Whether you worked for DELL or not, your statement about the Samsung drives is not a good one.
All manufacturers have issues at some point but to blanket a whole company rather than a range isnt good.

[Vadim]

i would blame vista, but thats me.
yet, if i were you, i would try to update the BIOS and intel drivers and see if you have the same problem

[marc0523]

Its very rare for hard drives to have intermittent problems.
They usually fail or they dont..
Whether you worked for DELL or not, your statement about the Samsung drives is not a good one.
All manufacturers have issues at some point but to blanket a whole company rather than a range isnt good.

Did Dell teach you to do that aswell?

Note the below post:

gagaga
Registered+

Join Date: May 2008
Location: London town
Posts: 19
Thanks: 0
Thanked 0 Times in 0 Posts

Re: Raid5 damaged .. no chance to rebuild ....
Originally Posted by valshare View Post

Quite probably - the Samsung drives are not suited to use in RAID arrays ... do a google on Samsung F1 RAID problems. They do make a specific RAID edition though.

The Samsung drives seem to go off and do their own thing for 20 or 30 seconds at time .. most RAID controllers give up on individual drives after about 10 seconds, then mark the drive as bad/unavailable.

Drives that are configured for use in RAID arrays generally give up trying to recover an unreadable sector and report a fail to the controller after 5-6 seconds ... the controller can then rebuild the data from the other drives, mark the original sector as bad and record the rebuilt data elsewhere. They also tend to have gone through a higher standard of qualification / quality control than normal drives. Desktop drives can try to recover data for 30 seconds, even minutes before giving up.

The cheapest way to get drives that behave well with RAID controllers is to go for WD drives - there is the WDTLER tool floating around the internet that tells standard (desktop) WD drives to report a fail quickly in the same way as the RAID editions do. Some people actually use the same tool to tell the higher-quality RAID editions to behave as desktop drives when used on their own ....

If the issue is with timing on the drives then this would be more likely to happen when transferring large amounts of data. I have only had my raid array fail when i have been transferring over 300gb of data.

Drives do not either Fail or Don't. Drives can slow down and die over time, SMART checking will tell you when your drive is near the end of its life because more sectors on the disk are coming back as bad.

I have had many problems with Samsung drives, i am going to be weary of them, that's called human nature.

Do a google search for "samsung raid dell xps fail site:support.dell.com"

Fixing several PCs a week for Samsung RAID issues and then having a RAID issue with Samsung drives I think keeping my distance from them in the future is a sensible policy.

And no, Dell did not teach me this.


Your whole reply was no help in resolving the problems with my systems RAID, it was just bringing up more issues.

Please reply with some helpfull information or ideas, not just trolling.

[leepen@scan]

Have you run the Samsung Drive Diagnostics to ensure that the disk is in good nick?

That's about the only thing I can think of that's worth checking, other than a rebuild of the array - unless you've already tried this?

[Blitzen]

Your whole reply was no help in resolving the problems with my systems RAID, it was just bringing up more issues.

Please reply with some helpfull information or ideas, not just trolling.

I dont 'troll'.
I was pointing out the fact to you that blaming a particualr drive, WITH NO EVIDENCE is just dense of the highest order.

If your original post contained any truth in your beliefs....or substance in your predictions i could almost undertand.
After you retort, and after my response to your quite frankly rude and stupid post, i dont care if your problems last a lifetime.

Just glad i dont buy Dell if your knowledge is anything to go by!

[Vadim]

Blitzen, your warm welcome to a new member of the forum is in my opinion a bit out of order.

This is a Scan support forum and any Scan customer has the right to ask for help in here, so if you have anything to contribute to solving the problem, you are welcome to do so, without insults though.

[marc0523]

Have you run the Samsung Drive Diagnostics to ensure that the disk is in good nick?

That's about the only thing I can think of that's worth checking, other than a rebuild of the array - unless you've already tried this?

Samsung Drive Diagnostics?

I'll check this out when I get home, I assume I can pick it up from their website?

[Philipp]

I have forwarded this to our repairs department

Very sorry to hear of the problems your having

[Ben @ Scan]

Hi the samsung drive test utility is called estool,please let us know if any of your drives fail and we can arrange an Rma and get the drive replaced

Cheers Ben

[marc0523]

OK, tried the Samsung diagnostics software and no errors came back.

I have killed my OS though now, BSOD with netio.sys.
RAID is missing.

Looks like i'll have to rebuild my os then...

[PeteG@Scan]

Hello,

Ok, a few comments on this thread overall.

We use the Samsung drives for a lot of systems and in a lot of configurations. As was mentioned it is rare for drives to have intermittent problems, and in those instances it's normally a outside issue causing it.

The whole Dell - samsung issue thing. Maybe you can google up a whole load of dell and samsung issues, but frankly that says more about the Dell boxes than it does about samsung drives.

I've worked on a few Dell boxes for friends recently. I noticed some pretty strange stuff whilst doing so. One of the p.c's had 2 drives in it. Both were assigned and softwared on the raid controller in a jbod configuration. If you removed one the other wouldn't boot.

I have never seen that anywhere else before.

The Dell bioses are written for dell. The raid config menu was also custom written, so honestly i'm not going to give the driver package written for the onboard controller on the custom made board the time of day.

The's a far higher chance that Dell simply can't get it's own hardware development together over a company like Samsung that churns out many many millions of units a year. We see thousands of these drives in system build here, and whilst I see many raid issues, they rarely if ever come down to a badly designed drive controller.

I don't have you build details here, so if you can post me your build number on here (or p.m. me your details) i'll look through it and see what stands out.

Something does strike me as being quite likely here, even more so now you've lost your netio.sys file.
I'll confirm that when I look over your build details.