[Search for users] [Overall Top Noters] [List of all Conferences] [Download this site]

Conference azur::mcc

Title:DECmcc user notes file. Does not replace IPMT.
Notice:Use IPMT for problems. Newsletter location in note 6187
Moderator:TAEC::BEROUD
Created:Mon Aug 21 1989
Last Modified:Wed Jun 04 1997
Last Successful Update:Fri Jun 06 1997
Number of topics:6497
Total number of notes:27359

5726.0. "Wildcarding hangs forms interface" by GIDDAY::MORETTI (Not drowning...waving) Tue Nov 09 1993 23:47

    Has anybody had any problems with wildcarding in the forms interface
    V1.3 ?
    
    I've got a customer that does the following :
    
    SHOW KNOWN BRIDGE * ALL
    
    and this hangs the interface, but if he puts in a valid bridge name all
    is well.
    
    He can also get the info from NCL using the wildcard command.
    
    Is there a fixkit or patch that addresses this problem as I can't find
    it listed in any of the release notes or patch notes on our local
    system ?
    
    Thanx
    
    John Moretti
    CSC Sydney
    Australia
    612-561-5109
T.RTitleUserPersonal
Name
DateLines
5726.1MCDOUG::dougImagine whirled peas.Wed Nov 10 1993 08:2517
>    
>    I've got a customer that does the following :
>    
>    SHOW KNOWN BRIDGE * ALL
>    

This is not a valid command, so I'd be *extremely* surprised if it even got
by the FCL parser.   

>    and this hangs the interface, but if he puts in a valid bridge name all
>    is well.

You may get more assistance it you can you post the EXACT, VERBATIM command
that your customer typed, along with other information like the system
configuration (OS, H/W, memory, disk, etc).

/doug
5726.2More infoGIDDAY::MORETTINot drowning...wavingThu Nov 11 1993 00:5122
    
    Sorry about the lack of info Doug but I thought it would be a known bug
    or something like that....
    
    Well the actual command the customer is using is :
    
    SHOW NODE BLAHBLAH LINK * ALL      or
    SHOW BRIDGE * ALL
    
    He is on a VAX4000/60 with 72M of MM and 2 x 1GIG disks. He states that
    if he substitutes the "*" with any valid name the interface is fine but
    wildcards seem to upset the FCL.
    
    He can use the wildcard in NCL with no problems but wildcards always
    hang the forms interface in any command that lends itself to wildcards.
    
    From other notes it appears that the wildcard entry should work....any
    clues ??
    
    Thanx again for your help
    
    John
5726.3now we're getting somewhere.TOOK::MCPHERSONImagine whirled peas.Thu Nov 11 1993 08:1529
>    SHOW NODE BLAHBLAH LINK * ALL      or
>    SHOW BRIDGE * ALL

    Ok.  Now we're getting somewhere.  ;^)

    Can you try using "ALL IDENT"  instead of just "ALL" ?    That command
    is ambiguous to the parser so I'd be surprised if it didn't just fail
    outright (i.e. it FCL parser can't tell if you mean "ALL IDENTIFIERS",
    "ALL STATUS", "ALL ATTRIBUTES" etc).   NOTE: If you just leave off the
    "ALL", FCL will default it to "ALL IDENTIFIERS".

>    He is on a VAX4000/60 with 72M of MM and 2 x 1GIG disks. He states that
>    if he substitutes the "*" with any valid name the interface is fine but
>    wildcards seem to upset the FCL.

    We need to figure out if it's going compute-bound or just _hanging_
    somewhere.  

    - Do you have a LOT of bridges registered?
    - Are you using DECdns or local MIR ?
    - If you do a $MON PROC/TOPCPU, does it look like the FCL is
      chewing up CPU ?

    If someone from the ELMS group will let you know the appropriate log
    bits to turn on (via MCC_BRIDGE_AM_LOG logicals), we should also be
    able to figure out what he Bridge AM is *doing* while the FCL is
    "upset".

    /doug
5726.4Bridge AM log bits...MOLAR::MOLAR::BRIENENNetwork Management Applications!Thu Nov 11 1993 09:3215
    /*************************************************************************/
    /* Bridge_log_type - Bits defs for mcc_bridge_am_log logical name        */
    /*                   (saved locally as VAR mcc_bridge_log)               */
    /*                                                                       */
    /* BRIDGE_k_DEBUG_ENTRY_POINTS (1) = breakpoints @ AM entry:DEBUG reqd!  */
    /* BRIDGE_k_TRACE_LOG (2h) = enable mcc_trace logging                    */
    /*  + BRIDGE_k_SCREEN_SWITCH (4h) = trace to screen or file              */
    /*  + BRIDGE_k_PERFORMANCE (10h)  = log performance data                 */
    /*  + BRIDGE_k_PACKET_DUMP (20h)  = dump select packets                  */
    /*                                                                       */
    /* Examples: 26(h) = output entry points and packet dumps to screen      */
    /*           22(h) = output entry points and packet dumps to file        */
    /*           03(h) = breakpoint at entry points + log entry pts to file  */
    /*************************************************************************/

5726.5Customer testing the DECNIS' s/wGIDDAY::MORETTINot drowning...wavingSun Nov 14 1993 19:3712
    
    Thanks for the good pointers, the customer is putting all this into
    action and has also found that if he does this command on a DECNIS500
    V2.0 it works OK but fails on DECNIS 600s with V2.0 and V2.2 .
    
    I am getting the customer to standardise the s/w on all DECNIS' and to
    set the BRIDGE_AM bits and to run the commands and fax the results back
    to me.
    
    Ciao
    
    John
5726.6No RBMS on DECNISMARVIN::COBBGraham R. Cobb (Routing), REO2-G/C2, 830-3917Thu Nov 18 1993 07:0912
The BRIDGE  AM implements RBMS which is not implemented on the DECNIS.  They
must  be  using  one  of  the other AMs in which case the BRIDGE_AM log bits
won't help!

Please get, from the customer, a precise description of exactly what command
they  issue  to  the  DECNIS and how it fails (I presume it hangs as that is
what  this  note  is  about!).  Also please ask them to try the same command
from  NCL, from MCC line mode and from the iconic map.  What system are they
running MCC on? What does the network look like between the MCC node and the
DECNIS (e.g. are they on the same LAN)?

Graham