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

Conference noted::sns

Title:POLYCENTER System Watchdog for VMS OSF/1 ULTRIX HP-UX AIX SunOS
Notice:Wishes:406,FAQ:845,Kits-VMS:1000,UNIX:694 VMS ECO01 FT kit: 521
Moderator:AZUR::HUREZZ
Created:Fri May 15 1992
Last Modified:Fri Jun 06 1997
Last Successful Update:Fri Jun 06 1997
Number of topics:1033
Total number of notes:4584

1018.0. " When is WDM event reported..." by CSC32::R_RIDGWAY () Mon Mar 17 1997 15:32

                           
    
    Hello;
    
    I thought I understood that the WDM event would be reported if
    the SNS agent was running when polled by a consoldator.  However
    I have noticed that if the SNS agent is stopped and the decnet
    object SNS$WATCHDOG has been declared, it reports;
    
             14-MAR 16:31 AXPBRD Connection status: connect to
                    network object rejected
    
    And when the SNS agent is not running the and the SNS$WATCHDOG
    object is not declared, it reports differently;
    
            14-MAR 16:44 AXPBRD Connection status: login information
                    invalid at remote node
    
    In the latter case, I tested this by manually stopping the agent
    and clearing the decnet object.
    
    I am working with PSW v2.2, using eco3 on both agent and consolidator,
    using phase IV decnet.  Once the SNS$WATCHDOG object is declared, I am 
    assuming the status returned back to the consolidator is a function of
    sys$system:sns$reject_once, if the agent is not running.
    
    I gratiously looking for a quick clarification if/when the 
    WDM event occurs.
    
    
    	Thanks;
    	 Rodger Ridgway
T.RTitleUserPersonal
Name		DateLines
1018.1Maybe is it obsolete in DECnet?AZUR::HUREZConnectivity & Computing Services @VBE. DTN 828-5159Tue Mar 18 1997 11:2050
    
    Looking into the code, the WDM event is reported when DECnet returns
    the condition value SS$_NOSUCHOBJ (documented below).  I couldn't get
    it actually generated indeed, even after stopping the Agent and renaming
    the command file sys$sysdevice:[sns$watchdog]sns$watchdog.com.  It seems
    that DECnet is always returning SS$_REJECT (connect to network object
    rejected) in that case indeed.  I'll try to investigate some more...
    
    Best Regards,
    
    	-- Olivier.
    
    ----------------------------------------------------------------------------
    $ help/message NOSUCHOBJ
    
    [...]
    
     NOSUCHOBJ,  network object is unknown at remote node
    
      Facility:     SYSTEM, System Services
    
      Explanation:  An attempt was made to initiate a DECnet for OpenVMS logical
                    link connection and the target object of the connect could
                    not be found on the specified node. The object name or
                    object number is unknown, or the command file used to start
                    the object could not be found.
    
      User Action:  Notify the system manager of the remote node. Refer to
                    the DECnet for OpenVMS Networking Manual for additional
                    information.
    ----------------------------------------------------------------------------
    $ help/message REJECT
    
    [...]
    
     REJECT,  connect to network object rejected
    
      Facility:     SYSTEM, System Services
    
      Explanation:  A request to connect to an object at a remote node
                    failed for one of the following reasons:
                    o The object requested does not exist.
                    o The access control information specified is invalid.
                    o The partner NSP has no resources.
                    o The partner task exited during the connect sequence.
                    o The partner task rejected the connect.
    
      User Action:  Determine which reason caused the failure and try to correct
                    the condition.
1018.2CSC32::BUTTERWORTHGun Control is a steady hand.Tue Mar 18 1997 16:5816
    Ollie,
      I was working on this with Roger and I distinctly remember getting
    the WDM event back in V1.0 days. I thought that the sns$reject_once 
    image that is run by SNS$WATCHDOG.COM should send a certain status code
    back to the consolidator and the consolidator would know that the
    Agent isn't running. The only time SNS$WATCHDOG.COM should run is when
    the agent isn't running due to the way that the SNS$WATCHDOG is
    defined in the permanent object database. Obviously, when the Agent
    starts and declares itself as the SNS$WATCHDOG object SNS$WATCHDOG.COM
    won't get executed. As I said earlier, I remember getting the WDM event 
    back in V1.0. Based on what you said in -1 about the NOSUCHOBJ status
    then it would seem that we'd only get the WDM event if the Agent was not 
    installed on a node we attempt to poll. Hmmmmm....
    
    Regs,
      Dan
1018.3AZUR::HUREZConnectivity & Computing Services @VBE. DTN 828-5159Wed Mar 19 1997 08:449
    
    I tried as well clearing the object definition in NCP, removing the
    SNS$WATCHDOG account, etc... with no chance!  The returned status
    always is connection rejected or refused.
    
    I still have to determine whether this is due to a change in SNS or in
    DECnet...
    
    	-- Olivier.
1018.4\CSC32::BUTTERWORTHGun Control is a steady hand.Wed Mar 19 1997 15:497
    Interesting and here's a new wrinkle. I just setup my V2.2-11
    consolidator node to poll a system that has never had Watchdog installed
    so you'd think we'd get a SS$_NOSUCHOBJ right? Wrong! We get 
    "login information invalid at remote node".
    
    Regs,
      Dan