Thanks.  That's about what I thought it might be.  (I have to do a "ps -ef | grep -v java" to get the effect of "ps -ef" on a Solaris/AIX system.)  Those older ones, as you indicate, may account for the nine zombies I have hanging around. 

The result would be to sum the Java process time and call it nvserverd.  That value makes sense to me. 

Anyone who can comment on the amount of resource used by the Switch Analyzer processes vis a vis the NetView ones? 

Bill Evans
Tivoli NetView support for DOE

-----Original Message-----
From: owner-nv-l AT lists.us.ibm DOT com [mailto:owner-nv-l AT lists.us.ibm DOT com] On Behalf Of James Shanks
Sent: Monday, June 27, 2005 5:09 PM
To: nv-l AT lists.us.ibm DOT com
Subject: Re: [nv-l] Java processes running

Bill,

It will take some experiments to determine whether what you see is normal
or not.
Basically the problem is that Linux treats java threads as separate
processes and displays them in ps -ef, whereas other UNIX OS's do not.
So if you are running the internal TEC adapter in nvserverd, for example,
you'll see six or seven additional java processes all started by nvserverd.
the trick is to trace things back through the PID and PPIDs.  That's about
all I can tell you, except that if any of your java processes  get parented
by init (1) and you start the NetView GUI with "netview",  then the code
will notice that and tell you that their are incorrectly parented (zombie)
processes being left behind.

James Shanks
Level 3 Support  for Tivoli NetView for UNIX and Windows
Tivoli Software / IBM Software Group