[CentOS] INIT: Id "snmp" respawning too fast: disabled for 5 minutes

Hello all,

I recently started seeing these messages on the consoles of three 
production Centos 5.2 servers. They have been occurring nonstop for the 
past few days and show up routinely every five minutes.

INIT: Id "snmp" respawning too fast: disabled for 5 minutes
INIT: Id "snmp" respawning too fast: disabled for 5 minutes
INIT: Id "snmp" respawning too fast: disabled for 5 minutes

Despite trying to debug this and googling for the last couple of hours, 
I haven't had any luck as to why the errors are generated or how to fix 
them. There are a ton of posts on the web related to other Id's, such as 
various tty ports and "x", but I haven't seen a single thing related to 
"snmp". The answer usually seems to be that there are misconfigured 
lines in /etc/inittab, but in my case there is nothing even remotely 
related to "Id snmp" in inittab.

The make matters more confusing, snmp isn't installed or being used on 
these machines. I doubled checked and the extent of any snmp packages 
are these two:

# rpm -qa | grep -i snmp
perl-Net-SNMP-5.2.0-1.2.el5.rf
net-snmp-libs-5.3.1-24.el5_2.1

Both only contain a couple of libraries, so I hardly believe they have 
anything to do with it.

According to what I've learned, this error is usually generated when 
init tries to continually respawn a process that dies immediately... but 
as to what process it could possible be I have no idea. I even tried to 
strace init (-p1), but apparently this is a forbidden operation:

# strace -p1
attach: ptrace(PTRACE_ATTACH, ...): Operation not permitted

That's a shame too, but at least I could see what init is trying to do.

Any ideas? I'm just about stumped as to how to proceed at this point.

Thanks in advance.

-- 
================================
David Halik
System Administrator
OIT-CSS Rutgers University
dhalik at jla.rutgers.edu
================================