[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

2615.0. ""No such entity" reported erroneously" by SHIPS::SMITH_J (Jeff Smith) Mon Mar 23 1992 16:15

    I have an urgent problem with DECmcc v1.1 on VMS, which will
    potentially loose us network management in a large UK bank.  We have
    DNS V1.1 on VMS, on a single node.
    
    The problem is that on a regular and cyclic basis, we are getting 
    "No such entity" being reported for entities that do exist, and which
    will be there the next minute.  This is seen at its clearest when
    trying to enable rules, where on moment a rule is there, and the next
    an equally valid rule is not there.  If the set of commands is re-run,
    (via @... at the MCC prompt), then maybe all rules will be OK, maybe
    others will be reported as "No such entity".
    
    Careful investigation shows that there is a clear 2 minute cycle to
    this behaviour, where every 2 minutes the "No such entity" message will
    be reported for 4/5 seconds, and then everything goes OK again.  It is
    as though something is being locked out, too busy, .... on these
    occasions, but we can't see any obvious jobs, processes or DECnet
    links.  On the later, it does seem that if the node is disconnected
    from the network, then no errors are reported - but then it is not much
    on a network management station!!  We believe that this error is 
    disabling rules in some way also.  
    
    This behaviour is giving rise to customer nervousness as to the
    stability of the product, and making it impossible to deploy on a VERY
    important testing / pseudo-production role.
    
    
    Any clues, suggestion to get to the bottom of it?
    
    Thanks,
    
    Jeff
T.RTitleUserPersonal
Name
DateLines
2615.1Need more informationMOLAR::ROBERTSKeith Roberts - DECmcc Toolkit TeamTue Mar 24 1992 10:3117
  Jeff,

  (Q) Exactly what are you (your customer) trying to do ?

      o What Rules are you creating (what is the Rule Expression) ?

  (Q) Which command is giving the 'no such entity' error message ?

      o Is it the Create or Enable Rule directive?
      o Or an error returned by Alarms when you do the
        'Show All Status' directive ?

  Any additional information you can provide will help us sort this out.

  Thanks,

  /keith
2615.2How many DNS servers?TEMTY::MINTZTue Mar 24 1992 10:465
    This is also one of the symptoms of problems with DNS.
    Is there more than one DNS server in the namespace you are using?
    
    -- Erik
    
2615.3One namespace, two competing implementations!!SHIPS::SMITH_JJeff SmithThu Mar 26 1992 10:4332
  I believe that I have got to the bottom of the problems being experienced at 
  Barclays with DECmcc.
  
    DECmcc is blameless, DECdns (Distributed Name Service) is giving rise
    to the  problem but is blameless.  The problem is that there are two
    DECdns  nameservers operating on an extended LAN  using the same
    namespace name (BARCLAYS:), but which have not been told  about each
    other.    
  
  The people who 'own' the DECdns namespace knew about one server, but did 
  not know about the other.  The  other people did know of the namespace 
  plans etc., but did not seem to understand what was required of them when 
  configuring their system.  
    
    I believe that MCC was using DNS quite happily.  Then (every 2
    minutes), the rogue DNS server advertised itself via the Ethernet, and
    caused great confusion, and MCC lost contact with the real DNS server
    momentarily.  At that time, I think it was linked to the rogue DNS
    server, and hence found that things that once were there, were now no
    longer registered.
    
    Bottom line:  The software did not have a problem, but there was a lack
    of  co-ordination over the network on the implementation of DECdns. 
    Even if you think you have an implemenation plan for DNS it is all too
    easy for someone to bring up another server and cause grief.
  
  Thanks to all who helped on this,
  
  Regards,
  
  Jeff