Anyone ever had an SATA cable cause data corruption? - Storage

This is a discussion on Anyone ever had an SATA cable cause data corruption? - Storage ; ASUS P5K Deluxe WiFi/AP 2GB PC6400 Giel RAM 1x 8800GTS 320MB 3x 500GB HDD Other drives 2x300GB and 1x320GB Is there a chance a SATA cable could be at the cause of data corruption? I've never heard of SATA cables ...

+ Reply to Thread
Results 1 to 5 of 5

Thread: Anyone ever had an SATA cable cause data corruption?

  1. Anyone ever had an SATA cable cause data corruption?

    ASUS P5K Deluxe WiFi/AP
    2GB PC6400 Giel RAM
    1x 8800GTS 320MB
    3x 500GB HDD
    Other drives 2x300GB and 1x320GB


    Is there a chance a SATA cable could be at the cause of data corruption?
    I've never heard of SATA cables causing problems, they either work or
    not in my experience. Let me explain...

    Currently got 3 SATA drives installed on the above system and am
    experiencing very occasional file corruption issues on whatever drive is
    connected to the 3rd port (tried 3 different drives) on there. The
    other 2 ports/drives seem to suffer no corruption.

    At first I thought it was maybe down to faulty RAM (I had to replace the
    RAM about a week or so before the corruption started occuring), but
    numerous long 48hr+ MEMEST86+ session give it the all clear.

    The corruption seems VERY subtle and mainly when large files are
    involved, it occurs (or atleast I can detect it) maybe once every 1 or 2
    days.

    I'm using the standard SATA cables that ASUS provide.

    Any ideas how to pin this down?



  2. Re: Anyone ever had an SATA cable cause data corruption?

    Zeno wrote:
    > ASUS P5K Deluxe WiFi/AP
    > 2GB PC6400 Giel RAM
    > 1x 8800GTS 320MB
    > 3x 500GB HDD
    > Other drives 2x300GB and 1x320GB
    >
    >
    > Is there a chance a SATA cable could be at the cause of data
    > corruption? I've never heard of SATA cables causing problems, they
    > either work or not in my experience. Let me explain...
    >
    > Currently got 3 SATA drives installed on the above system and am
    > experiencing very occasional file corruption issues on whatever drive
    > is connected to the 3rd port (tried 3 different drives) on there. The
    > other 2 ports/drives seem to suffer no corruption.
    >
    > At first I thought it was maybe down to faulty RAM (I had to replace
    > the RAM about a week or so before the corruption started occuring),
    > but numerous long 48hr+ MEMEST86+ session give it the all clear.


    > The corruption seems VERY subtle and mainly when large files are involved,


    What do you mean by that exactly ? Is that only seen when you
    copy large files or is it also seen when you read just them, and is the
    corruption permanent when it happens or variable from read to read ?

    > it occurs (or atleast I can detect it) maybe once every 1 or 2 days.


    How are you deciding that you are seeing corruption ?

    > I'm using the standard SATA cables that ASUS provide.


    > Any ideas how to pin this down?


    The obvious way to do that is to automate a CRC etc of the file etc.



  3. Re: Anyone ever had an SATA cable cause data corruption?

    Previously Zeno wrote:
    > ASUS P5K Deluxe WiFi/AP
    > 2GB PC6400 Giel RAM
    > 1x 8800GTS 320MB
    > 3x 500GB HDD
    > Other drives 2x300GB and 1x320GB



    > Is there a chance a SATA cable could be at the cause of data corruption?


    Not really.

    > I've never heard of SATA cables causing problems, they either work or
    > not in my experience. Let me explain...


    > Currently got 3 SATA drives installed on the above system and am
    > experiencing very occasional file corruption issues on whatever drive is
    > connected to the 3rd port (tried 3 different drives) on there. The
    > other 2 ports/drives seem to suffer no corruption.


    > At first I thought it was maybe down to faulty RAM (I had to replace the
    > RAM about a week or so before the corruption started occuring), but
    > numerous long 48hr+ MEMEST86+ session give it the all clear.


    I have had bit-errors in RAMs that memtest86+ did not find.
    I had to run an unserspace memterster under Linux....

    > The corruption seems VERY subtle and mainly when large files are
    > involved, it occurs (or atleast I can detect it) maybe once every 1 or 2
    > days.


    Can you determine the exact nature of the corruption, i.e. can
    you compare a corrupted file with an uncorrupted version?

    > I'm using the standard SATA cables that ASUS provide.


    > Any ideas how to pin this down?


    First, SATA uses checksums ion all transfers, uncaught corruption
    is extremely unlikely.

    The suspects are:

    - RAM (memtest86+ seems npt to do a good job anymore today, unfortunately)
    - CPU (bus/cache)
    - System bus
    - Chipset, especially SATA controller

    Since you have corruption only on one port, I would say the most
    likely candidate (almost certain) is the SATA controller. All other
    candidates should cause corruption on the other disks as well. It is
    possible thet the SATA controller causes some corruption on its 3rd
    port internally, before the data gets fitted with checksums.

    Advice: RMA the board.

    Arno



  4. Re: Anyone ever had an SATA cable cause data corruption?

    Zeno wrote:

    > ASUS P5K Deluxe WiFi/AP
    > 2GB PC6400 Giel RAM
    > 1x 8800GTS 320MB
    > 3x 500GB HDD
    > Other drives 2x300GB and 1x320GB
    >
    >
    > Is there a chance a SATA cable could be at the cause of data corruption?
    > I've never heard of SATA cables causing problems, they either work or
    > not in my experience. Let me explain...
    >
    > Currently got 3 SATA drives installed on the above system and am
    > experiencing very occasional file corruption issues on whatever drive is
    > connected to the 3rd port (tried 3 different drives) on there. The
    > other 2 ports/drives seem to suffer no corruption.
    >
    > At first I thought it was maybe down to faulty RAM (I had to replace the
    > RAM about a week or so before the corruption started occuring), but
    > numerous long 48hr+ MEMEST86+ session give it the all clear.
    >
    > The corruption seems VERY subtle and mainly when large files are
    > involved, it occurs (or atleast I can detect it) maybe once every 1 or 2
    > days.
    >
    > I'm using the standard SATA cables that ASUS provide.
    >
    > Any ideas how to pin this down?
    >
    >


    Since those cables are cheap and easy to swap, I suggest replacing both
    the signal and power cables to that HD to see if the problem goes away.
    Like Arno, I think the other hardware is more likely than cables, but
    cables are easier to test.

    The SATA signal cable is even less likely to be the culprit than the
    SATA power cable; that low-tech power cable doesn't even have CRC. ;-)
    --
    Cheers, Bob

  5. Re: Anyone ever had an SATA cable cause data corruption?

    On Thu, 03 Apr 2008 20:50:45 +0100, Zeno put finger to
    keyboard and composed:

    >ASUS P5K Deluxe WiFi/AP
    >2GB PC6400 Giel RAM
    >1x 8800GTS 320MB
    >3x 500GB HDD
    >Other drives 2x300GB and 1x320GB
    >
    >
    >Is there a chance a SATA cable could be at the cause of data corruption?
    > I've never heard of SATA cables causing problems, they either work or
    >not in my experience. Let me explain...
    >
    >Currently got 3 SATA drives installed on the above system and am
    >experiencing very occasional file corruption issues on whatever drive is
    >connected to the 3rd port (tried 3 different drives) on there. The
    >other 2 ports/drives seem to suffer no corruption.
    >
    >At first I thought it was maybe down to faulty RAM (I had to replace the
    >RAM about a week or so before the corruption started occuring), but
    >numerous long 48hr+ MEMEST86+ session give it the all clear.
    >
    >The corruption seems VERY subtle and mainly when large files are
    >involved, it occurs (or atleast I can detect it) maybe once every 1 or 2
    >days.
    >
    >I'm using the standard SATA cables that ASUS provide.
    >
    >Any ideas how to pin this down?


    Compare your good and bad files. You can do this from a CMD window by
    typing ...

    fc /b good_file bad_file

    Consistent differences in a particular bit could point to an error in
    some data bus. A system RAM error could also produce the same symptom,
    but I expect this would affect all your ports/drives.

    - Franc Zabkar
    --
    Please remove one 'i' from my address when replying by email.

+ Reply to Thread