poor performance after sata installed drive

Discussion in 'Abit' started by Martyn Clarke, Dec 13, 2003.

  1. I currently have the below spec.

    nf7-s
    3000+ amd
    80gb ibm (master)
    60gb ibm (master)

    after install of another drive as a master on sata 1 there is a marked
    reduction of boot up speed from approx 25-45 secs to nearly 2 mins. i have
    updated driver, bios and anything else that i can think of. am i missing
    something obvious or is this fixed by another process.

    any comments please.

    thanks martyn
     
    Martyn Clarke, Dec 13, 2003
    #1
    1. Advertisements

  2. Martyn Clarke

    boohoo Guest

    I have this problem on my kx7-333r.....when i installed a HD to ata133 port
    1, then at the highpoint screen there is a pause for 30
    seconds......unconnect no pause.....

    anyone have any suggestions why is this....or it just is if u attach a hd to
    ata133?

    ta
     
    boohoo, Dec 14, 2003
    #2
    1. Advertisements

  3. Martyn Clarke

    Bill Drake Guest

    Hi, Martyn. The delay is the result of a search for a drive that is
    no longer connected. In this situation, the system will wait the
    entire length of a drive-detect-timeout -- before continuing the
    boot process.

    The normal drive-detect-timeout for PC Systems is somewhere
    between 30sec and 180sec per drive (the actual number varies
    according to the BIOS designer's intent).


    So, what you need to do is figure out where the "mortal remains"
    of the configuration for the "phantom" drive(s) is/are located, and
    then remove the offending configuration-item(s).


    Things to check:

    1. Ensure you have no CMOS entries corresponding to drives
    that no longer exist. Yes this includes floppy, Zip, CD, DVD,
    and Hard Disks on both the motherboard-chipset controllers
    and the external controllers.

    Anything the CMOS says is present -- which is not *actually*
    present -- will induce a drive-detect delay.


    2. Ensure that each Hard Disk connected to a SATA channel
    is connected as a Master Drive in *single drive* mode.

    With native-SATA drives, this is usually automatic and there
    is no possibility of misconfiguration. PATA drives, connected
    to a SATA channel through a Serillel2 adaptor are *quite* a
    different matter.

    The jumpers on a PATA drive must be confirmed as correct
    for Master/Single operation when that drive is used on a
    SATA channel.

    Note: With some manufacturers (Eg: Quantum/Maxtor) the
    drive is set to Master/Single automatically if the drive is
    jumpered as a Master drive.

    Other manufacturers (Eg: Western Digital) have a
    *specific* jumper setup for Master/Single -- which is
    *different* from the jumper setup for Master with
    Slave-Present.

    Leaving a drive that has a Master/Slave-Present jumper
    setup in place -- with that drive connected to a SATA
    channel -- will *induce* a drive-detect-timeout.

    The PATA drive will dutifully tell the SATA controller
    (through the Serillel2) to wait for that missing "slave"
    drive to come ready. After the drive-detect-timeout,
    the PATA drive will finally give up on waiting for the
    missing "slave" drive (that the jumpers dutifully tell the
    "master" is supposed to be there) and the SATA
    controller will then work with the existing "master"
    drive (and quite correctly so).

    Each drive connected to a SATA channel must be
    configured as a Master/Single device. Any other
    configuration will induce drive-detect delays.



    3. Ensure that you have no "signature" entries on any *single* Hard
    Disk connected to a controller that can possibly act as a RAID
    device.

    Note: It is very common for some RAID-BIOS versions to
    automatically create a "signature" for *both* drives when
    two drives are present -- even if those two drives are
    not configured to be used in an array.

    Then, if one or the other drives is removed, the "signature"
    for the remaining drive tells the RAID-BIOS there is
    supposed to be another drive present.

    Consequently, the RAID-BIOS waits the entire duration of
    the Hard Disk drive-detect-timeout for that other drive (that
    the "signature" says is supposed to be there) to come-ready.

    Only after the drive-detect-timeout has expired does the
    RAID-BIOS allow the boot process to continue -- using the
    one remaining drive in standalone mode.


    If a "signature" on a single drive is present in error, use the
    RAID-BIOS capacity to *remove* the signature from any drive
    that runs as a standalone device.

    (Now you know why that mysterious entry in the RAID-BIOS is
    present in the first place. It is there *precisely* to handle the
    very situation described above.)



    In the SIL SATA RAID-BIOS Configuration Screens, the item
    used to solve this problem is called "Resolve Conflicts". SIL
    call the RAID "signature" on the drive its metadata.


    For more info, see the NF7-S Manual entries in Chapter 5,
    section 5.2.4. Further info can also be found the Help file
    for the SIL SATA-RAID Configuration Utility, as per the NF7-S
    Manual entries under Appendix D.


    Note to lurkers: In the Intel ICH5R RAID-BIOS, the entry
    described above as "Resolve Conflicts" is
    called "Reset Disks to Non-RAID".

    This item is present in the ICH5R RAID-BIOS
    for *exactly* the same reasons it is present
    in the SIL RAID-BIOS.

    ICH5R Users need to be aware of this
    possibility in their own setups as well.



    Best I can do for now. <tm>


    Bill
     
    Bill Drake, Dec 14, 2003
    #3
    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.