Thanks to:
Nikola Milutin Tony Gale Marc Potvin
Merlin Wilkerson Bennet Fauber Thomas Leitner
Roger Leonard Oisin McGuiness Stephan Spalding
Sean O'Connell John P Speno
I got about four suggestions total from the above list of respondents. The
overwhelming favorite was to place a "nohup" at the beginning of the command
line which starts the process. This is the one that worked for me.
As for the other suggestions:
No, I'm not using an "su".
No, there is no PID file multiple instanced of this daemon are allowed.
Yes, it is written in "sh" not "ksh" or any other shell.
One of the respondents mentioned that he found recommendations to try
"nohup" and "sleep" commands in a summary in the archives. Obviously I
didn't find that summary. If I add SUMMARY to my archive search and I get a
large number of unrelated questions stating the obvious, "I'll summarize"
which is a requirement of the list anyway, so I'll stop saying it. (Yes, one
of my own questions was returned, with that phrase.)
Here, is my original question for anyone interested:
Hello Managers,
I have an interesting problem, I have a startup script called ups in
/sbin/init.d and is linked to S99ups in /sbin/rc3.d,(as well as
/sbin/rc0.d/K01ups).
The problem is this, when I run:
# /sbin/rc3.d/S99ups start
The process starts up with out any problems. When I reboot, it doesn't, no
error messages or anything, the process doesn't start up, it just isn't
there.
I changed the first line of the startup script to read:
#!/bin/sh -x
^^
So I could see what was happening and what's going wrong. The script seems
to run equally well either from the command line or during the reboot. I see
the same output, it tells me it is up and running. But after the reboot it's
simply not there.
Anybody had a similar problem? As always, I'll summarize.
Jim Fitzmaurice
jpfitz_at_fnal.gov
UNIX is very user friendly, It's just very particular about who it makes
friends with.
Received on Tue Nov 02 1999 - 17:53:08 NZDT