[Intel Navigation Header]

    LANDesk(R) Management Suite v2.x: DMI/Miscellaneous Usage Tips

    Contents:

    • CONTENTS
    • HOW TO TURN A WORKSTATION INTO A DMI MANAGED DEVICE
    • REBUILD THE CORE DATABASE
    • "UNABLE TO ALLOCATE HANDLE FOR IMMEDIATE DISCOVERY CYCLE"
    • "NETWORK DATABASE ERROR - THE DISCOVERED DATA MAY BE INACCURATE"
    
    

    CONTENTS

    
       This document discusses issues commonly encountered with DMI (Desktop
    
       Management Interface) or the core database.  It supplements the files in
    
       LANDesk(R) Management Suite's Readme Viewer.
    
    
    
    

    HOW TO TURN A WORKSTATION INTO A DMI MANAGED DEVICE

    
       Every workstation that you want to manage with the DMI must load the
    
       service layer provided with LANDesk(R) Management Suite v2.0. This
    
       service layer is a 26 KB TSR called SL.EXE.  If you are using Windows*
    
       on a workstation, the service layer for Windows must ALSO be loaded. A
    
       MIF (Management Information File) can be created for each workstation by
    
       running LDISCAN /M on the workstation. By default the MIF file is stored
    
       on the workstation hard drive in the DMI/MIF directory.
    
    
    
    

    REBUILD THE CORE DATABASE

    
       Starved Btrieve* environment space may cause your LANDesk(R) Management
    
       Suite databases to become corrupt. Another issue could be that your file
    
       server Cache Buffers have dropped below 40 percent. If this happens, you
    
       need to either unload other NLMs or add system memory. The core
    
       management server requires at least 12 MB of system RAM, with 4 MB
    
       dedicated to LANDesk Management Suite's NLMs. This recommendation is
    
       running Netware* (no NDS) and only LANDesk Management Suite on the
    
       server. In most environmnets a minimum of 24MB of system RAM is
    
       required. Managed Servers in your LANDesk Management Suite domain
    
       require 550 KB of system RAM.
    
    
    
       Corrupt database messages or symptoms indicate a logical corruption of
    
       the file, missing data records, etc., but the file is physically intact.
    
       In some instances, a file is physically damaged so badly that Btrieve is
    
       unable to process it. 
    
    
    
       Some error messages you may see indicating a physically of logically
    
       corrupt database are:
    
    
    
       "Database Error: This is an undefined error" 
    
       "Btrieve Error 86: creating new database" 
    
        "DISCMAN: The network database is corrupted - discovered data may be
    
       inaccurate"
    
    
    
       In all cases you 'll want to increase the Btrieve environment space and
    
       rebuild the database.
    
    
    
       If you are suspicious of the integrity of the database, use the steps
    
       below to refresh the database (order is very important). To troubleshoot
    
       why the database corruption occurred in the first place, see the Btrieve
    
       Issues on the pink card included in the shipping release documentation.
    
    
    
            1. Unload NWDSDISC (NetWare v4.02 or v4.1 only)
    
            2. Unload DISCMAN
    
            3. Unload PHYSMAP
    
            4. Load PHYSMAP NEWDB
    
            5. Load DISCMAN VERBOSE
    
    
    
            6. Load NWDSDISC (NetWare v4.02 or v4.1 only)
    
    
    
       Wait a minute and try launching the LANDesk Management Console again.
    
       You probably don't want to modify your LD_AUTO.NCF to load PHYSMAP.NLM
    
       and DISCMAN.NLM with these parameters. The verbose option is optional.
    
       See document 5615 if AMS.NLM seems to have any trouble with the AMS
    
       database (Btrieve causing high CPU utilization); you may need to rebuild
    
       the AMS database by reloading AMS with the NEWDB option.
    
    
    
       If database corruption continues, the most likely cause is starved
    
       Btrieve resources. Utilize Btrieve's BTRMON.NLM utility in isolation
    
       this potential cuase. IN BTRMON, if you select Active Resources, you
    
       will see a list of open files. You can press the <Insert> key to update
    
       the displayed list. The other thing that BTRMON can show you is the
    
       resource utilizaiton - both current and peak. This screen should tell
    
       you right away what resource you are running out of. 
    
    
    
    

    "UNABLE TO ALLOCATE HANDLE FOR IMMEDIATE DISCOVERY CYCLE"

    
       If DISCMAN.NLM generates the message "Unable to allocate handle for
    
       immediate discovery cycle," the LANDesk(R) Management Suite database may
    
       have become corrupt due to insufficient Btrieve* environment space, or
    
       the file server has below 40 percent cache buffers free. To rebuild the
    
       database, see Rebuild the core database. To make sure Btrieve is not
    
       starved for environment space, see the Btrieve Issues on the pink card
    
       in the shipping release documentation.
    
    
    
    

    "NETWORK DATABASE ERROR - THE DISCOVERED DATA MAY BE INACCURATE"

    
       If the message "Network Database error - the discovered data may be
    
       inaccurate" appears at the file server console, the LANDesk(R)
    
       Management Suite database may have become corrupt due to insufficient
    
       Btrieve* environment space, or the file server has below 40 percent
    
       cache buffers free, in most situations it is best to run with
    
       approximately 60 percent cache buffers free.. To rebuild the database,
    
       see Rebuild the core database. To make sure Btrieve is not starved for
    
       environment space, see the Btrieve Issues on the pink card in the
    
       shipping release documentation.
    
    

    Trademark information