Wanda, I have seen this, sort of, and it is expected behavior, working as 
designed.

When you run expire with duration, or use the 'cancel expiration' command as I 
do,
the point where expire stops is recorded in the db.  When you start it again
it starts at that point and continues either for the duration or until the 
logical
end of the database is reached.  If it reaches the end of the db, then expire
considers that it has completed 1 pass over the database and quits.

When expire was improved, the code of course had a bug in it; hitting the 
duration
did not store the stop point.  The 'cancel exp' command did store the stop point
which is why I use the command and not the duration parameter.  At some
point the bug was fixed and that is probably when you started noticing the
change in behavior.

I run expire weekly, starting at 5 am, and a cancel expire command is issued at
11 pm.  The next week expire starts at 5 am but usually finishes around noon.

Hope this helps,


--
--------------------------
--------------------------
Bill Colwell
Bill Colwell
C. S. Draper Lab
Cambridge, Ma.
bcolwell AT draper DOT com
--------------------------
In <1D06C9567A6AD4119391000629D9009D975ED8 AT central.jhuapl DOT edu>, on 
05/30/01
In <1D06C9567A6AD4119391000629D9009D975ED8 AT central.jhuapl DOT edu>, on 
05/30/01
   at 05:11 PM, "Prather, Wanda" <Wanda.Prather AT JHUAPL DOT EDU> said:

>TSM 3.7.4 on AIX 4.3.3

>I have EXPIRE INVENTORY scheduled to run for 1.5 hours in the wee hours:
>EXPIRE INVENTORY QUIET=YES DURATION=90


>Sometimes EXPIRE INVENTORY runs OK.
>Other times it starts up, examines/expires 200-300 objects, then shuts
>itself down again in 2-3 minutes with the normal message of SUCCESS.

>Now I know it should be expiring 100,000+ objects per day.
>If I start it again manually, it takes off and runs as you would expect.

>Does anyone else see this happen?
>Any idea what causes the premature shutdown?