Please report all spam threads, posts and suspicious members. We receive spam notifications and will take immediate action!
Page 1 of 4 123 ... LastLast
Results 1 to 10 of 32

Thread: GA-EX58-UD5 Raid5 unreliable




  1. #1
    Join Date
    May 2009
    Posts
    32

    Default GA-EX58-UD5 Raid5 unreliable

    A few months ago I built a new system with:
    GA-EX58-UD5 Rev: 1.0
    CPU Intel Core i7 920
    Mem Corsair TR3X6G1600C8D (6GB total)
    HSF Noctua NH-U12P SE1366
    PSU Corsair CMPSU-620HX
    VGA XFX HD 5770
    HDD four (4) 2TB Hitachi HDS722020ALA330
    Lite-On iHOS104-08 Blu-ray BDROM (port 4)
    HDD (system) Seagate 1.5TB ST31500341AS (port 5)
    Win7-64 Ultimate

    The four Hitachi drives are connected to ports 0-3, RAID, native command mode, Raid5, GPT.
    The system drive is on port 5.

    On three separate occasions several weeks apart, the system reported that one of the RAID5 drives failed. The first two times, it was the same drive on port 1. On each of those occasions I ran Hitachi's extended drive test and NO problems were found (ALL drives were long tested prior to system build). On both occasions I attempted to recover the array by resyncing that drive but it reported it would take 70 hours so I just deleted the array and recreated and formated after copying off the data to a temporary location. After the second incident, I swaped the drives on ports 0 and 1.

    Yesterday for the third time now the RAID5 became degraded again. This time it claimed the drive on port 3 failed. At that time, I was running BIOS F10. I upgrade to BIOS F12 to see if that alone cleared the problem but of course it didn't. I didn't bother a long test this time and just tried to recovery the array. Even though only 668GB of 5.45TB was used it was taking so incredibly long I again decided to copy off the data to a temporary location, delete, create, format, and copy back the data.

    The system is still spread out on top of my desk in pieces because every time I'm about to put it in its case the RAID array fails and I figure I need to resolve the issue before going to all the work of stuffing this thing into the tower. It is possible that not being in a case is causing problems (because of grounding issues, static, movement, etc.)?

    Oh, and one other oddity, when using Hitachi's bootable disc diagnostic CD, I have to test the drives on the slower gigabyte ports because the Hitachi diagnostic program does NOT see any drives on the Intel ICHR10 ports (0-5).

    Do I have a flaky disc controller?

    Intel(R) Matrix Storage Manager report below.

    Suggestions?

    Thank you.

    668GB of 5.45TB used
    System Information
    Kit Installed: 8.9.0.1023
    Kit Install History: 8.9.0.1023, Uninstall
    Shell Version: 8.9.0.1023
    OS Name: Microsoft Windows 7 Ultimate
    OS Version: 6.1.7600 Build 7600
    System Name: NEW
    System Manufacturer: Gigabyte Technology Co., Ltd.
    System Model: EX58-UD5
    Processor: Intel(R) Core(TM) i7 CPU 920 @ 2.67GHz
    BIOS Version/Date: Award Software International, Inc. F10 , 02/06/2010
    Language: ENU

    Intel(R) Matrix Storage Manager
    Intel RAID Controller: Intel(R) ICH8R/ICH9R/ICH10R/DO/PCH SATA RAID Controller
    Number of Serial ATA ports: 6

    RAID Option ROM Version: 8.0.0.1039
    Driver Version: 8.9.0.1023
    RAID Plug-In Version: 8.9.0.1023
    Language Resource Version of the RAID Plug-In: 8.9.0.1023
    Create Volume Wizard Version: 8.9.0.1023
    Language Resource Version of the Create Volume Wizard: 8.9.0.1023
    Create Volume from Existing Hard Drive Wizard Version: 8.9.0.1023
    Language Resource Version of the Create Volume from Existing Hard Drive Wizard: 8.9.0.1023
    Modify Volume Wizard Version: 8.9.0.1023
    Language Resource Version of the Modify Volume Wizard: 8.9.0.1023
    Delete Volume Wizard Version: 8.9.0.1023
    Language Resource Version of the Delete Volume Wizard: 8.9.0.1023
    ISDI Library Version: 8.9.0.1023
    Event Monitor User Notification Tool Version: 8.9.0.1023
    Language Resource Version of the Event Monitor User Notification Tool: 8.9.0.1023
    Event Monitor Version: 8.9.0.1023

    Array_0000
    Status: No active migrations
    Hard Drive Data Cache Enabled: Yes
    Size: 7452 GB
    Free Space: 0 GB
    Number of Hard Drives: 4
    Hard Drive Member 1: Hitachi HDS722020ALA330
    Hard Drive Member 2: Hitachi HDS722020ALA330
    Hard Drive Member 3: Hitachi HDS722020ALA330
    Hard Drive Member 4: Hitachi HDS722020ALA330
    Number of Volumes: 1
    Volume Member 1: raid5

    raid5
    Status: Degraded
    System Volume: No
    Volume Write-Back Cache Enabled: No
    RAID Level: RAID 5 (striping with parity)
    Strip Size: 64 KB
    Size: 5589 GB
    Physical Sector Size: 512 Bytes
    Logical Sector Size: 512 Bytes
    Number of Hard Drives: 4
    Hard Drive Member 1: Hitachi HDS722020ALA330
    Hard Drive Member 2: Hitachi HDS722020ALA330
    Hard Drive Member 3: Hitachi HDS722020ALA330
    Hard Drive Member 4: Hitachi HDS722020ALA330
    Parent Array: Array_0000

    Hard Drive 0
    Usage: Array member
    Status: Normal
    Device Port: 0
    Device Port Location: Internal
    Current Serial ATA Transfer Mode: Generation 2
    Model: Hitachi HDS722020ALA330
    Firmware: JKAOA20N
    Native Command Queuing Support: Yes
    Hard Drive Data Cache Enabled: Yes
    Size: 1863 GB
    Physical Sector Size: 512 Bytes
    Logical Sector Size: 512 Bytes
    Number of Volumes: 1
    Volume Member 1: raid5
    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: Hitachi HDS722020ALA330
    Firmware: JKAOA20N
    Native Command Queuing Support: Yes
    Hard Drive Data Cache Enabled: Yes
    Size: 1863 GB
    Physical Sector Size: 512 Bytes
    Logical Sector Size: 512 Bytes
    Number of Volumes: 1
    Volume Member 1: raid5
    Parent Array: Array_0000

    Hard Drive 2
    Usage: Array member
    Status: Normal
    Device Port: 2
    Device Port Location: Internal
    Current Serial ATA Transfer Mode: Generation 2
    Model: Hitachi HDS722020ALA330
    Firmware: JKAOA20N
    Native Command Queuing Support: Yes
    Hard Drive Data Cache Enabled: Yes
    Size: 1863 GB
    Physical Sector Size: 512 Bytes
    Logical Sector Size: 512 Bytes
    Number of Volumes: 1
    Volume Member 1: raid5
    Parent Array: Array_0000

    Hard Drive 3
    Usage: Array member
    Status: Failed
    Device Port: 3
    Device Port Location: Internal
    Current Serial ATA Transfer Mode: Generation 2
    Model: Hitachi HDS722020ALA330
    Native Command Queuing Support: Yes
    Hard Drive Data Cache Enabled: Yes
    Size: 1863 GB
    Physical Sector Size: 512 Bytes
    Logical Sector Size: 512 Bytes
    Number of Volumes: 1
    Volume Member 1: raid5
    Parent Array: Array_0000

    Hard Drive 4
    Usage: Non-RAID hard drive
    Status: Normal
    Device Port: 5
    Device Port Location: Internal
    Current Serial ATA Transfer Mode: Generation 2
    Model: ST31500341AS
    Firmware: CC1H
    Native Command Queuing Support: Yes
    System Hard Drive: Yes
    Size: 1397.2 GB
    Physical Sector Size: 512 Bytes
    Logical Sector Size: 512 Bytes

    CD/DVD Drive 0
    Device Port: 4
    Device Port Location: Internal
    Current Serial ATA Transfer Mode: Generation 1
    Model: ATAPI iHOS104
    Firmware: WL0B

  2. #2

    Default Re: GA-EX58-UD5 Raid5 unreliable

    Quote Originally Posted by gigabyter View Post
    Oh, and one other oddity, when using Hitachi's bootable disc diagnostic CD, I have to test the drives on the slower gigabyte ports because the Hitachi diagnostic program does NOT see any drives on the Intel ICHR10 ports (0-5).
    Not sure, but this might be similar to:
    I did got a X25-M SSD that lacked latest firmware. Intel flasher naturally recommend's ICH ports (0-3), and AHCI mode. Sure, device on port 4 or 5 (GA-EP45-DQ6) was not seen, and on ports 0-3 stalled on detection (RAID, Native IDE). Only when reverting temporarily to plain AHCI mode did the utility spot the SSD. So how well tools can reach the "raw iron" seem to depend on the mode of the controller.


    "When IDE drive (SATA or PATA) notices a bad sector, it remaps logical sector to one in reserve. The remapping takes a while. The SCSI drive, once considered "smart" compared to IDE, simply tells the controller that there is a problem. RAID controllers, used to responsive SCSI drives, think that busy (remapping) IDE is dead, and mark it off." This at least was the standard explanation earlier, and Western Digital has used that in their marketing, boasting with "Raid Edition" drives that do respond within timeout.

    But it sounds odd that Hitachi drives would not behave. At least they should show "reallocated_sectors" in their S.M.A.R.T. data.

    Overall, RAID5 does require checksum computations, unlike raid modes 0 and 1. Sure, the CPU is mighty these days, but still a real hardware chip with cache memory is likely to fare better than the ICH fake raid.

  3. #3
    Join Date
    Nov 2008
    Posts
    25

    Default Re: GA-EX58-UD5 Raid5 unreliable

    Hi,

    Nice system. I have a similar setup that is a few years old now with a RAID5 array on 4 samsung SpinPoint T166, 500 Go (OS on separate drive). At first this was on an Asus P5W DH Deluxe (ICH7R) but later moved to a Gigabyte GA-EP45-DS3R (ICH10R). On both boards I have nothing but problem after problem with RAID disc drops. Ive upgraded bios, intel matrix RAID drivers, changed cables, turned off the write cache etc all to no avail. Some times the system will be stable for 6 months but then it will drop a disc a couple of times in the same week. Sometimes the drop is in the OS (vista) - sometimes it comes in BIOS matrix RAID manager after a BSOD. With 500 Gb discs the reabuild time is 24 hours - with 2T that multiplies out to a LONG time...
    Frankly after my several years of trying to get my RAID5 stable I don't like your chances of getting a system you can rely on. One thing I havent tried yet is to swap out the power supply - I was just in the process of doing this (replacing the Zalman ZM600-HP with a Corsair CMPSU-620HX (your PSU as it happens) when my mobo died (http://forums.tweaktown.com/f69/ga-e...63/#post349122). The fact that you have this same issue with the PSU I wanted to try doesn't fill me with hope... Next time I do RAID (if ever) the 2 things I'll do are:
    1. Spend the extra and go specialist RAID drives (most brands have a dedicated RAID line of drives that cost about 30% more - these drives don't do to error correction which can cause write delays and hence drops - apparantly).
    2. Get a proper RAID controller card (eg "+getMessage("iPrintVerKit")+"). The on board intel chipsets RAID "controllers" are only "FakeRAID" ie processing is offloaded onto CPU - they dont have the dedicated hardware controller these cards have . It would be interesting to try your current drives in the RocketRAID hardware controller - that might fix your problems - Id be really interested to know - if not sell the drives and replace with RAID drives plus controller - in my view that is going to be the only way your going to sleep at night.
    Something else I havent tried is to drop down to RAID1 mirror with just 2 drives - run that for 6 months or so and see if there are issue - maybe without the parity checksum computation of RAID5 there will be less issues? (I know this is not a long tem solution as you'l want more than 2Tb).
    Also its worth reading this article about large disc RAID5 arrays (Why RAID 5 stops working in 2009 | Storage Bits | ZDNet.com) - you might want to think about spending a little more on the RAID controller card and getting something that does RAID6?
    I'm really interested in knowing if you get this sorted - so please post back anything you try that works. But what ever you do expect a LONG road a head - after various changes my system was sometimes great for 6 months or so before several rapid failures... you have to wait 12 months or so to really be sure you are clear. Good luck.
    Last edited by Ricibob; 03-29-2010 at 07:21 PM.

  4. #4
    Lsdmeasap's Avatar
    Lsdmeasap is offline GIGABYTE Guru
    Join Date
    Jan 2008
    Location
    PCB Island
    Posts
    25,940

    Default Re: GA-EX58-UD5 Raid5 unreliable

    This can often happen with unstable settings, are you guys 100% stable? Fully tested with Prime95 or Orthos and Memtest86+?

  5. #5
    Join Date
    Nov 2008
    Posts
    25

    Default Re: GA-EX58-UD5 Raid5 unreliable

    My system is just a HTPC (+very light gaming) and so is not OCed in any way. Early on I ran memtest86 overnight a couple of times with no errors - I cant remember if I really stress tested it at all - I think I might have early on (when it had the Asus mobo) just to see how the case temps went - and I dont remember any issues. Having said that the system is NOT stable - ie it BSODs quite sporadically and I have never really had the time to investigate fully the cause. Some of them I am pretty sure are video card driver related as they happen when MediaCenter probs card at video start up (stop code 0x100000ea - apparantly often video related) but some others seemed to be directly RAID related ( 0x000000d1 DRIVER_IRQL_NOTLESS_OR_EQUAL - iaStor.sys). The iastor.sys driver is related to Intel RAID (quite a few people arround with this issue) and Intel have issued a fix in form of Matrix RAID drivers - I installed those recently and was really hopeful about this putting an end to my problems but since then have seen a spontanious dropped disk in windows and when I restarted the machine for some reseason (don't know if related or not) it wouldn't even POST (LSD you're familiar with this story... http://forums.tweaktown.com/f69/ga-e...63/#post349173). Anyway it probably wouldnt do gigabyter any harm to update to the latest Matrix drivers if he hasn't already.

  6. #6
    Lsdmeasap's Avatar
    Lsdmeasap is offline GIGABYTE Guru
    Join Date
    Jan 2008
    Location
    PCB Island
    Posts
    25,940

    Default Re: GA-EX58-UD5 Raid5 unreliable

    Ahh, well those tests are not enough, you need to run Prime95, Orthos, ect. << To the above users mainly

    Ya I know of our discussions, I just replied to your thread now. Nice to see you are here right now while I am. This way I can reply to you quickly so if need be you can try the BIOS thing we are discussing.

  7. #7
    Join Date
    May 2009
    Posts
    32

    Default Re: GA-EX58-UD5 Raid5 unreliable

    Thanks for all the replies.

    The system is not OC'd, however am using profile 1 for the memory.

    Fully tested with Memtest86+. Will test with Prime95 64-bit.

    I guess I'm spoiled by my current system based on ASUS A8N32-SLI. I've been running this system with it's Raid5 for four years without a signle drive drop (4 drive 1TB Raid5).

  8. #8
    Join Date
    Nov 2008
    Posts
    25

    Default Re: GA-EX58-UD5 Raid5 unreliable

    Wow 4 years without a drive drop - Im blown away... If I had a dollar for every drive drop my system has been through I could practically buy a new one.... Prior to the Gigabyte board I originally had an Asus P5W DH Deluxe (Intel ICH RAID like the Gigabyte) and I had exactly the same RAID5 issues with it (plus other Vista wake up from sleep issues) - maybe its the Nvidia chipset on the ASUS A8N32-SLI that makes the difference. You would think that Intel would know how to put together a chipset to manage RAID5 it looking like that just isnt the case. (Those 4 x 1Tb disks must have cost a bit of cash 4 years ago...)

  9. #9
    Join Date
    May 2009
    Posts
    32

    Default Re: GA-EX58-UD5 Raid5 unreliable

    Quote Originally Posted by Ricibob View Post
    Wow 4 years without a drive drop - Im blown away... If I had a dollar for every drive drop my system has been through I could practically buy a new one.... Prior to the Gigabyte board I originally had an Asus P5W DH Deluxe (Intel ICH RAID like the Gigabyte) and I had exactly the same RAID5 issues with it (plus other Vista wake up from sleep issues) - maybe its the Nvidia chipset on the ASUS A8N32-SLI that makes the difference. You would think that Intel would know how to put together a chipset to manage RAID5 it looking like that just isnt the case. (Those 4 x 1Tb disks must have cost a bit of cash 4 years ago...)
    Actually the RAID5 array size is 1TB. It' has benn so long that I can't remember the individual drives but they might be 320GB. I think I'm using the on-board Silicon Image SiL 3132 SATALink Controller for the RAID5 array.

    I wonder if it might be worth using the Gigabyte RAID controller instead.

  10. #10
    Join Date
    Nov 2008
    Posts
    25

    Default Re: GA-EX58-UD5 Raid5 unreliable

    I wonder if it might be worth using the Gigabyte RAID controller instead.
    Yep would have been a good option (anything would better than ICH10R) - except it looks like it doesn't do RAID5 - Gigabyte page home page says "GIGABYTE SATA2 chip: 4 x SATA 3Gb/s connectors Support for SATA RAID 0, RAID 1 and JBOD". Because each drop your going to be looking down the barrel of 3-4 days solid rebuilt - I think a RocketRAID controller card is going to be your only solution - but maybe Im just pesimistic. When I get my dead board backup up thats the route Im going - Ive had enough ICH10R for one lifetime.

Page 1 of 4 123 ... LastLast

Thread Information

Users Browsing this Thread

There are currently 1 users browsing this thread. (0 members and 1 guests)

Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts
  •