Conference iosg::all-in-1_v30

    Ann,
    
    I don't understand the last paragraph about starting the EAST server
    on SOUTH... but for the initial problem does SET WATCH show anything.
    
    Regards,
    
    Andrew.D.Wicks

    i did a SET WATCH and don't see any problems (ie: no error messages).
    
    when i try to start the SERVER for EAST i get the following
    
    	on EAST, get the message shareable images must be installed to run
    privileged image 
    
    	on SOUTH or NORTH, the job gets submitted to EAST$BATCH, but the
    log file shows the following error..
    
    %RUN-S-PROC_ID, identification of created process is 216002C0
    Error writing to server process mailbox - %RMS-E-FNF, file not found
    %DCL-W-UNDFIL, file has not been opened by DCL - check logical name
    %DCL-W-UNDFIL, file has not been opened by DCL - check logical name
    %NONAME-W-NOMSG, Message number 00000000
    Server is being started:
                    configuration_file:
    DSA2004:[A1DATA.DATA_SHARE]EAST$SERVER73.DA
                    process_name:       EAST$SRV73
    
    there is a DSN article with this error, it says the cause is the way
    that the symbol SYSUAF is defined. all 3 nodes of this cluster are
    using the same SYSUAF and it is defined the same on all 3 nodes.
    
    is everyone else confused, too?
    
    	ann
    
    welcome back andrew

    Ann,
    
    and what does the SYSUAF logical look like 
    does object 73 exists in NCP
    why do you have 3 FCSs and not just one for the whole cluster?
    
    note 1083 also shows a match.
    
    Regards,
    
    Andrew.D.wicks

    Customer had the same problem the other day.
    Mine was because A1$batch was removed after some works
    over the weekend.
    
    Not too sure if yours is caused by that.
    
    ching-U

    
    >%RUN-S-PROC_ID, identification of created process is 216002C0
    >Error writing to server process mailbox - %RMS-E-FNF, file not found
    >%DCL-W-UNDFIL, file has not been opened by DCL - check logical name
    >%DCL-W-UNDFIL, file has not been opened by DCL - check logical name
    >%NONAME-W-NOMSG, Message number 00000000
    >Server is being started:
    >                configuration_file:
    >DSA2004:[A1DATA.DATA_SHARE]EAST$SERVER73.DA
    >                process_name:       EAST$SRV73
    
    If this is EXACTLY what the system says, then check the server config
    file name (R in the MS menu). There is a missing T on the file name.
    Could be a typo though.  The message you are seeing looks to be 
    caused by the startup procedure tyring to pass the config file into
    the server process, the server process can't find the file.
    
    --Bob
    
    

    re .5
    
    the missing T is just a "typo". i did a cut and paste from 1 window to
    another and the T was off screen on the cutting window.  i did however
    check the MS R and the proper name for the configuration file is there.
    
    Configuration:Type:               LOCAL                                       
                  Process Name:       EAST$SRV73                                  
                  Startup State:      ENABLED                                     
                  Startup Queue:      EAST$BATCH                                  
                  Configuration File: OA$DATA_SHARE:EAST$SERVER73.DAT             
    re .3
    
    OBJECT_73 is NOt on node EAST, but is on NORHT and SOUTH. i checked
    with the VMS system manager and this is a voltile NCP object and is
    created upon startup of ALL-IN-1.
    where exactly does OBJ_73 get created?
    
    Ann
    

    
    Ann
    
    Object 73 is the default decnet object number for the FCS. it is 
    created as part of the fcs startup.
    
    Can you try the following and see what happens:
    
    $ fcs :== $sys$system:oafc$server.exe
    $ fcs oa$data_share:east$server73.dat
    
    This will try and start the FCS up interactively. You should either
    get a message saying that server east::"73=" has started, or some
    sort of error message saying why it didn't. If it starts up, then 
    you should stop it before moving to the next step (control-y works
    fastest here).
    
    If that works, the next step (we are working from the bottom up
    trying to find the problem here), is to start the FCS detached using
    the fcs startup procedure:
    
    $ @sys$startup:oafc$startup oa$data_share:east$server73.dat
    
    This one you will have to wait a few minutes for it to start, then
    do either a $show proc/id=x (where x is given as a result of the 
    above line), or do an $mcr ncp sho obj obj_73.
    
    Either should show you the FCS running. If it is running have someone
    do something that connects to it (index of a non-MAIN drawer) just 
    to be sure.
    
    To stop this you should do a stop $ proc/id=x. You probably can manage
    this one from withing IOS as well.
    
    if this works, then the problem is either the IOS interface, or the
    communication between the IOS UI and the FCS startup procedure.
    
    --Bob
    

    re .7
    
    thanks bob!
    
    sep #1 gave me a different error message. 
    
    EAST_mgr> fcs :== $sys$system:oafc$server.exe
    EAST_mgr> fcs oa$data_share:east$server73.dat
    %DCL-W-ACTIMAGE, error activating image OAFC$MTS_PRIV_SHR
    -CLI-E-IMGNAME, image file
    DSA2001:[SYS2.SYSCOMMON.][SYSLIB]OAFC$MTS_PRIV_SHR.EX
    E;2
    -SYSTEM-F-PROTINSTALL, protected images must be installed
    
    I looked on NORHT/SOUTH and the image is there. i didn't see it on EAST
    did an INSTAL CREATE on EAST, loggeed out, ran the above test, got the
    same error.
    
    ann - "installs r us"

    
    Make sure it is installed the same way on EAST as on the others.
    
    Also, if this fails, check oa$build for a more recent version and
    try this one (this image is built during the link of the FCS and
    should be the correct one).
    
    If that fails then, ???????
    
    --Bob