Conference azur::mcc

    Dan, It would not be too hard to include your code fragment in
    DNA4 AM. However, we would need to make some modifications to
    make it work in the CMA threads environment we exist in. We
    would need to run a verification for memory leaks to make sure
    we didn't break something. 
    
    I think this is an excellent suggestion for post v1.2. I expect
    that you are going to be frustrated with that answer. Let me
    explain. We are freezing all code changes to DNA4 AM on Feb. 14.
    The people who are doing DNA4 AM changes are working 10+ hour
    days to get everything that is currently scheduled done. We also
    need to minimize the risk of last minute changes for any
    component breaking the systems builds. There is no way I can
    justify adding additional work 2 weeks before a functional
    freeze. 
    
    I can consider doing further work after the freeze and putting it
    in a seperate CMS branch and once V1.2 has been shipped rolling
    the new branch into the main line. For a specific customer (ie.
    yours), we can then have a special DNA4 AM that represents
    your functionality increase. 
    
    That accomplishes your immediate need. Once we have fully tested
    the change in the customer site, we can make the upgraded DNA4 AM
    available internal to DEC for additional testing, etc.
    
    If this is an acceptable solution please contact me via mail and
    lets work out the details.
    
    wally

  For v1.2 the toolkit has prepared the Example FM which implements the
  PONG directive.  Simply put, the Example FM calls the Sample AM's 
  SHOW IDENT directive and times the response.  If it takes too long you
  get an event.  The code is pretty generic.

  I would like to see an architected directive called PING (or maybe theres
  a better name).  This directive would be generic and function pretty much
  as I described for the Example FM.  Every global entity would get this
  directive for free -- that is they don't have to write any special code to
  get it.  Their Show Identifiers directive would 'tickle' their entity.
  The call could be timed out, resulting in a timeout-event; entity dead.

  Now -- if you have a better technology specific way of determining if your
  entity is alive, then *you* implement the PING directive yourself.  You just
  have to follow the guidelines of the PING directive.

  /keith

/* -------------------------------------------------------------------------- */
/* ////////////////////////// Simple Check ////////////////////////////////// */
/* -------------------------------------------------------------------------- */
	Hi -
		I'd like a stab at this. The DNA4-AM currently has code to do
this work. We could just remove those sections needed and take Keith's 
suggestion and create a ping directive. The steps are as follows straight from
the AM.
		( 1 ) Build a network control block
                ( 2 ) Connect to the entity
                     ( a ) Check the various status if not normal
                     ( b ) Check to see if this is a phaseiv or phasev node
                ( 3 ) update our table accordingly

                /* ----------------------------- */
                /* //// We could stop here //// */
                /* ----------------------------- */

                ---> OR MORE <---

                ( 4 ) send our message which is a SHOW IDENT to the NODE4
                ( 5 ) Read our reply from that NODE4
                ( 6 ) update our table accordingly






- darryl
/* -------------------------------------------------------------------------- */
/* ////////////////////////////////////////////////////////////////////////// */
/* -------------------------------------------------------------------------- */