B5151 Hardware Failure

Discussion in 'Tyan' started by David M. Mroz, Mar 23, 2006.

  1. I have a B5151 that spontaneously throws a blue screen in Windows Server
    2003. The message simply says: "Hardware Failure. Please contact your
    vendor." and that's it. I called Tyan Tech support and they're trying to
    tell me that it *could* be the RAM. We switched it out to no avail.

    I could replicate the problem instantly by pulling out one of the
    hotswappable drives. We spent a day replacing each of the drives an
    rebuilding the arrays thinking it might be one of the drivers. It seemed
    stable and stayed up for about a week, but last night it crashed again.

    TSM says the temperatures might be a little high, but in the safe range.
    No entries in the system log that point to anything in specific.

    Any ideas of what this could be?

    Setup:
    Intel Pentium 4 3.0 GHz Processor
    2.0 GB RAM (2x1GB)
    4 Western Digital SATA Caviar RE Drives in RAID-1 configuration (2x2)
    Windows Small Business Server 2003 (32-bit)
    I can get more detailed specs about each of the components if needed.

    Thanks

    Dave
     
    David M. Mroz, Mar 23, 2006
    #1
    1. Advertisements

  2. David M. Mroz

    R. J. Salvi Guest

    Is this a new problem, or has it happened since day 1? Any new hardware
    additions?

    A few things to try:

    * can you/have you tried removing the top cover to eliminate temps as the
    culprit? CPU heatsink maybe get jarred accidentally while R&R'ing drives?
    * try the RAM in slots 2&4 instead of 1&3
    * do the SATA drive ribbons correspond to their respective places on the
    backplane board and BIOS settings? i.e. SATA1 to SATA1, SATA2 to SATA2, etc.
    * have you tried removing half the mirror array? 1x2 instead of 2x2...try
    both 1x2s
    * Intel chipset drivers installed?

    Regarding your mouse wandering...are you using the mfrs' drivers? When you
    switched from Logitech to MS, did you uninstall the Logitech software?

    If both problems persist (mouse & BSOD), it might be time to RMA the board.

    Hmmm...
     
    R. J. Salvi, Mar 23, 2006
    #2
    1. Advertisements

  3. <snippety-two>

    Just some preliminary notes. First, I would never trust Caviar-
    class drives in a server application, nor would I depend on anything
    other than SCSI-based RAID-5. There are good reasons that big servers,
    like the Dell PowerEdge and Compaq Proliant series, depend on such in
    their own hardware.

    Now, with that said: My first suspect in a failure like this would
    be the power supply. You'd be surprised how often the power side of any
    system is neglected, or cut corners with.

    The only way to know for sure, regrettably (for most), is to pull
    the supply out of the system and bench-test it with a variable
    electronic load and an oscilloscope to check for ripple and noise while
    it is under said load.

    If your existing supply is not made by PC Power and Cooling
    (http://www.pcpowercooling.com), I would look at swapping it out for one
    that is. You can also check with Tyan Support to see which supply they
    recommend.

    Happy hunting.
     
    Dr. Anton T. Squeegee, Mar 24, 2006
    #3
  4. Thanks for your reply - I appreciate the suggestions.

    It's a problem that started recently - it ran fine for a couple of months
    and then all of the sudden started crapping out on us. Sometimes it'll run
    for a week, other times for an hour - it's difficult to predict when it'll
    go down. For instance, we reconfigured it on Wednesday and it ran until
    Sunday at 6:00am and we rebooted it at Noon. It crapped out at 3:00 am and
    we rebooted it at 4:30, and then it died again at 5:30 am. So there seems
    to be no real pattern.

    We didn't try running it without the cover. The software utility says that
    it's running at 55C on the processor. I've seen some things that seem to
    suggest that is too high, and others that say it's right in range.

    We replaced the RAM and the SATA backplane to eliminate those as options.
    And we checked the cables to ensure that they match. I can see that they do
    in the case, in the BIOS and in the Intel Utility that comes with it. And
    yep, the Intel drivers are installed. I asked Tyan if there were any
    updates to the drivers or BIOS and they said no. (Both system BIOS and RAID
    BIOS).

    I don't have a Logitech mouse - just standard PS2 mouse. I was just saying
    that I see my mouse flip out every once in awhile.

    Thanks again for your help!

    Dave
     
    David M. Mroz, Mar 27, 2006
    #4
  5. This application didn't need the cost of SCSI RAID-5 - it's a barebones
    server that started as SATA and that's what we decided to use. We've built
    plenty of servers from the ground-up based on SATA RAID architecture and
    they're all running very well. It figures that the one time that we try
    increase reliabilty by buying a pre-built server that things get worse and
    reliabiltiy goes to zero.

    Aside from the physical processor and motherboard, the power supply is the
    one component that hasn't been replaced. It may be the next thing that we
    try before RMAing the system.

    Dave
     
    David M. Mroz, Mar 27, 2006
    #5
    1. Advertisements

Ask a Question

Want to reply to this thread or ask your own question?

You'll need to choose a username for the site, which only take a couple of moments (here). After that, you can post your question and our members will help you out.