Ghost Solution Suite

 View Only
Expand all | Collapse all

Ghost Client Error

  • 1.  Ghost Client Error

    Posted Jan 11, 2007 01:16 PM
    Can someone provide me with some info on what this means or why Ghost Client is crashing on statup or when starting of the service?


    The following information has been placed on the clipboard. Visit the Ghost forums at http://forums.symantec.com/discussions/forum.jspa?forumID=109 and supply this information to gain help with this error.
    Symantec Ghost Client Agent
    An exception has occurred of type C0000005
    C:\Program Files\Symantec\Ghost - C:\Program Files\Symantec\Ghost\ngctw32.exe 11.0.0.1502

    e:\depot\ghost\gsstrunk\ghost\configureagent\client\getuuid.cpp#18:l_checksum
    @3 unsigned int len = 0x3511
    e:\depot\ghost\gsstrunk\ghost\configureagent\client\getuuid.cpp#153:g_getUuid
    bool sysValid = false
    bool dmiValid = false
    e:\depot\ghost\gsstrunk\ghost\configureagent\client\clientmain.cpp#3149:ServerStatus::formatStatus
    e:\depot\ghost\gsstrunk\ghost\configureagent\client\clientmain.cpp#3169:ServerStatus::buildStatus
    unsigned int idle = 0x42FC81
    unsigned int ready = 0x71AB2BC0
    unsigned int busy = 0xAC32A13
    e:\depot\ghost\gsstrunk\ghost\configureagent\client\clientmain.cpp#3252:ServerStatus::doFoundServer
    e:\depot\ghost\gsstrunk\ghost\configureagent\client\locator.cpp#331:Locator::gotMessage
    e:\depot\ghost\gsstrunk\ghost\configureagent\agentlib\sockchange.cpp#633:DatagramInterchange::gotMessage
    0x7270E
    e:\depot\ghost\gsstrunk\ghost\configureagent\agentlib\sockchange.cpp#541:DatagramInterchange::readEvent
    0x63742


  • 2.  RE: Ghost Client Error

    Posted Jan 11, 2007 04:49 PM
    What it's trying to do is obtain the machine's unique BIOS ID, which is the preferred method of identifying clients in GSS2 (before that it used the network adapter MAC address), but accessing this BIOS data from Windows is a little tricky up until Windows 2003 which added an API for it.

    In fact, there are two standards which define this ID, and two methods for getting it; one is defined by the System Management BIOS specification, and one is defined by the PXE specification. The code is in the middle of trying to decode the PXE one, but it seems to have hit a problem decoding the PXE bios data.

    The first thing to look at is how your machine's BIOS formats this block of BIOS data. I generally find the simplest way to do this is with the good old DOS debug command. Open a command window on the client, and enter this:
    C:\> debug

    You can get out of the debugger back to the command line using the command "q". In the debugger, enter this search command:

    -s f000:0 ffff "_SYSID_"

    That command asks the debugged to seach the BIOS data segment for the string _SYSID_ which is one of the "eye-catchers" that marks the presence of this block. For instance, in this notebook I'm using this spits out

    F000:7550

    If it doesn't print anything, then the string wasn't found so it might be in an alternative area which you can seach using

    -s e000:0 ffff "_SYSID_"

    Next, I take that number that it printed and ask it to print the memory at that location

    -d f000:7550
    F000:7550 5F 53 59 53 49 44 5F BF-11 00 61 75 0F 00 01 00 _SYSID_...au....
    F000:7560 00 5F 55 55 49 44 5F C4-19 00 44 45 4C 4C 51 00 ._UUID_...DELLQ.
    F000:7570 10 4C 80 37 CA C0 4F 4C-31 53 00 00 00 00 00 00 .L.7..OL1S......
    F000:7580 5F 53 4D 5F 80 1F 02 04-FD 00 00 00 00 00 00 00 _SM_............
    F000:7590 5F 44 4D 49 5F 10 CC 09-A0 75 0F 00 3B 00 24 90 _DMI_....u..;.$.
    F000:75A0 DA FB 00 DA B2 00 0D 5F-0F 37 40 7D 00 00 00 00 ......._.7@}....
    F000:75B0 00 7E 00 02 00 00 00 40-00 04 00 01 00 41 00 04 .~.....@.....A..
    F000:75C0 00 00 00 90 00 05 00 00-00 91 00 05 00 01 00 92 ................


    That block of data contains a string "_UUID_" which is the eye-catcher for the PXE UUID. In my system, this is fine and the code has no trouble with it - your system appears to have this data block formatted unusually, and I have to figure out how to tweak the code to cope with it.

    At this point, if you can cut-and-paste this data out of the command window and into a text file (command windows have a system menu which allows you to mark an area of the text window - like any normal Windows window, alt-space is a keyboard shortcut which opens the system menu) and either post it here or e-mail it to me at nigel dot bree at gmail dot com to analyse.


  • 3.  RE: Ghost Client Error

    Posted Jan 11, 2007 05:31 PM
    Here is the info that it returned that you requested

    Microsoft Windows XP
    (C) Copyright 1985-2001 Microsoft Corp.

    C:\>debug
    -s f000:0 ffff "_SYSID_"
    F000:6AB0
    -df000:7550
    F000:7550 80 F1 51 80 F1 50 88 00-51 C1 03 50 89 00 50 8A ..Q..P..Q..P..P.
    F000:7560 00 51 8E 03 51 8F 03 51-6F 02 51 76 02 51 77 02 .Q..Q..Qo.Qv.Qw.
    F000:7570 51 7E 02 51 7F 02 51 80-02 51 81 02 51 82 02 51 Q~.Q..Q..Q..Q..Q
    F000:7580 83 02 00 F8 01 50 F8 10-50 FC 02 50 18 11 50 8B .....P..P..P..P.
    F000:7590 00 51 10 F2 51 00 F2 50-8E 00 00 48 13 50 4A 13 .Q..Q..P...H.PJ.
    F000:75A0 01 04 30 01 88 F0 01 20-F2 01 98 70 01 30 72 01 ..0.... ...p.0r.
    F000:75B0 38 72 50 F2 00 51 FE 00-51 FA 10 51 FF 00 01 F0 8rP..Q..Q..Q....
    F000:75C0 73 01 00 21 01 03 01 01-04 01 01 05 01 01 06 01 s..!............
    -


  • 4.  RE: Ghost Client Error

    Posted Jan 11, 2007 06:00 PM
    Thanks, but I was a little unclear in my instructions - you need to ask it to dump the memory based on the number that the "s" command prints. Since it printed f000:6AB0 for you, you need to put that number in the "d" command, like this:

    C:\>debug
    -s f000:0 ffff "_SYSID_"
    F000:6AB0
    -d f000:6AB0


  • 5.  RE: Ghost Client Error

    Posted Jan 12, 2007 10:23 AM
    Heh yea i goofed that up here is the new info.

    Microsoft Windows XP
    (C) Copyright 1985-2001 Microsoft Corp.

    C:\>debug
    -s f000:0 ffff "_SYSID_"
    F000:6AB0
    -d f000:6AB0
    F000:6AB0 5F 53 59 53 49 44 5F 69-11 00 C1 6A 0F 00 02 00 _SYSID_i...j....
    F000:6AC0 00 5F 55 55 49 44 5F 5D-19 A0 32 22 64 DF 1D B2 ._UUID_]..2"d...
    F000:6AD0 11 80 00 9C 00 C7 56 D6-6F 00 5F 31 33 39 34 5F ......V.o._1394_
    F000:6AE0 49 11 A0 32 22 64 DF 1D-B2 11 00 00 00 00 00 00 I..2"d..........
    F000:6AF0 24 46 53 58 60 0C ED 00-B2 00 E0 00 00 00 00 00 $FSX`...........
    F000:6B00 5F 33 32 5F 20 D6 0F 00-00 01 D7 00 00 00 00 00 _32_ ...........
    F000:6B10 24 50 44 4D 01 0B 04 74-87 00 F0 00 00 00 00 00 $PDM...t........
    F000:6B20 24 48 42 44 0C CD 01 00-D5 8D 00 F0 DE 00 00 00 $HBD............


  • 6.  RE: Ghost Client Error

    Posted Jan 14, 2007 04:29 AM
    Thanks for that. It seems that the BIOS manufacturer for your clients unfortunately got the conformance with the PXE specification wrong - it seems that they are out by a byte in the location of the UUID, which is what is messing up my code.

    I'm creating some modifications to the parsing code for this structure which should make it work with this situation; drop me a line at nigel dot bree at gmail dot com if you want to help test out a development build of an updated client that deals with this.


  • 7.  RE: Ghost Client Error

    Posted Mar 09, 2007 10:56 AM
    Nigel,
    I have just upgraded my Client Console to 2.0. I have several Gateway tablets which appear to be experiencing this issue. Is there a fix yet?
    Thanks,
    Chris O.

    Here are the debug results:

    U:\>debug
    -s f000:0 ffff"_SYSID_"
    F000:6720
    -d f000:6720
    F000:6720 5F 53 59 53 49 44 5F FC-11 00 31 67 0F 00 02 00 _SYSID_...1g....
    F000:6730 00 5F 55 55 49 44 5F 11-19 EB 04 2E FF 54 01 5E ._UUID_......T.^
    F000:6740 58 C3 33 C9 C3 50 53 52-43 00 5F 31 33 39 34 5F X.3..PSRC._1394_
    F000:6750 60 11 00 00 00 00 00 00-00 00 00 00 00 00 00 00 `...............
    F000:6760 24 46 53 58 56 0C F7 00-B2 00 E0 00 00 00 00 00 $FSXV...........
    F000:6770 5F 33 32 5F 60 D7 0F 00-00 01 96 00 00 00 00 00 _32_`...........
    F000:6780 00 00 00 00 01 10 48 3B-4E A2 6E 00 00 00 00 00 ......H;N.n.....
    F000:6790 24 50 44 4D 01 0B 75 04-86 00 F0 00 00 00 00 00 $PDM..u.........
    -Message was edited by:
    Chris Oestereich


  • 8.  RE: Ghost Client Error

    Posted Mar 20, 2007 03:23 PM
    Nigel,
    I have just upgraded my Client Console to 2.0. I have several Gateway M405 notebooks which appear to be experiencing this issue. Is there a fix yet?
    Thanks,
    David McDaniel

    C:\>debug
    -s f000:0 ffff"_SYSID_"
    F000:6330
    -d f000:7550
    F000:7550 00 00 00 10 00 F6 03 00-00 01 00 00 00 04 00 00 ................
    F000:7560 00 14 00 00 40 00 00 00-00 00 00 07 01 00 00 3C ....@..........<
    F000:7570 00 FF FF FF FF 70 01 00-00 08 00 00 00 04 00 00 .....p..........
    F000:7580 00 18 00 76 03 00 00 01-00 00 00 04 00 00 00 1C ...v............
    F000:7590 00 00 80 00 00 00 00 00-00 07 01 00 00 3C 00 FF .............<..
    F000:75A0 FF FF FF F0 03 00 00 06-00 00 00 04 00 00 00 00 ................
    F000:75B0 00 F7 03 00 00 01 00 00-00 04 00 00 00 00 00 04 ................
    F000:75C0 00 00 00 00 00 00 00 08-00 00 00 00 00 40 00 00 .............@..


  • 9.  RE: Ghost Client Error

    Posted Mar 20, 2007 06:18 PM
    -s f000:0 ffff"_SYSID_"
    F000:6330
    -d f000:6330
    F000:6330 5F 53 59 53 49 44 5F F0-11 00 41 63 0F 00 02 00 _SYSID_...Ac....
    F000:6340 00 5F 55 55 49 44 5F B8-19 EB 18 2E FF 54 01 5F ._UUID_......T._
    F000:6350 72 11 2E 8A 17 E8 1C FD-03 00 5F 31 33 39 34 5F r........._1394_
    F000:6360 60 11 00 00 00 00 00 00-00 00 00 00 00 00 00 00 `...............
    F000:6370 52 53 44 20 50 54 52 20-F1 47 41 54 45 57 41 00 RSD PTR .GATEWA.
    F000:6380 40 6A 6E 1F 00 00 00 00-00 00 00 00 00 00 00 00 @jn.............
    F000:6390 00 00 00 00 00 00 00 00-00 00 00 00 00 00 00 00 ................
    F000:63A0 24 46 53 58 56 0C F7 00-B2 00 E0 00 00 00 00 00 $FSXV...........


  • 10.  RE: Ghost Client Error

    Posted Mar 20, 2007 08:49 PM
    Oops, sorry I missed Chris's message (this forum software isn't the best so it's hard to keep track of old threads getting bumped).

    Yes there's a fix, just e-mail me as above to get it.