Most likely you have a script issuing snmptrap somewhere that is causing 
this problem.  Do you know whether you have some of these?
In versions 7.1.x, often the problem is that the trap variables are larger 
than 512 bytes, a restriction that was enforced as part of the CERT fixes 
awhile back.

But the way to see the invalid trap is turn on the hex dump of all packets 
for trapd, using serversetup.  This causes him to run with the -x option. 
Once trapd runs with -x, then you can toggle the trapd trace on and off 
using the  command "trapd -T".   The trace will capture the trap in hex 
and you can match up the trace with the time stamps in trapd.log to see 
exactly what trap this is.  You may find it difficult to interpret the hex 
data, though, so I'd recommend you get Support involved.  They will be 
happy to decode the trap for you and then you can determine how it was 
generated.

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




"CATALINA MARTINEZ" <CATALINA.MARTINEZ AT tlc.state.tx DOT us>
01/07/2003 09:51 AM

 
        To:     <nv-l AT lists.tivoli DOT com>
        cc: 
        Subject:        [nv-l] invalid SNMPTrap



Netview 712 AIX 433 ML10
 
I am getting the following trap:
1041946888 7  Tue Jan 07 07:41:28 2003 <none>  T   WARNING: invalid 
SNMPTrap packet from agent 10.24.0.100 source A pid -1
 
the agent/address is the Netview Server.. 
 
What does this mean? How do I fix it?
 
Thanks