snmpwalk and snmpget are two different operations. In the first case you
don't have to specify a starting variable (because we provide a default for
you), and if you choose to do so, then you can specify a branch node rather
than a leaf node of the MIB tree, because snmpwalk will just keep asking
for the next leaf of the tree until none are left (snmpwalk does repeated
SNMP getnext calls.). But for snmpget you must specify the exact MIB
instance, the leaf, you want or you'll get an error back. Table elements
have an index as an instance number. Simple objects have an instance
number of zero, by convention.
Thus, to get sysUpTime returned from snmpwalk you can say
snmpwalk .1.3.6.1.2.1.1.3 <node name>
For snmpget you must specify the "0" instance
snmpget .1.3.6.1.2.1.1.3.0 <node name>
Your ruleset is going to do snmpget, so you have to specify the instance
you want there too.
What you should get back is just the numeric value. The textual stuff you
see on the command line when you run the commands there is added by that
executable to make the result more understandable.
I have no explanation about your sniffer results.
James Shanks
Level 3 Support for Tivoli NetView for UNIX and Windows
Tivoli Software / IBM Software Group
"Glen Warn"
<Glen.Warn@pemcoc
orp.com> To
Sent by: <nv-l@lists.us.ibm.com>
owner-nv-l@lists. cc
us.ibm.com
Subject
RE: [nv-l] Ruleset ? using Compare
04/28/2005 02:04 MIB Variable
PM
Please respond to
nv-l
Hi James,
I have much more info now. I do have several good debugs saying the
problem is "CompareMIBDnode.C[159] : SNMP GET operation for this MIB
object failed." This was really helpful. Since then, I've tried the
numeric OID .1.3.6.1.2.1.1.3 and the worded syntax
.iso.org.dod.internet.mgmt.mib-2.system.sysUpTime. Neither seem to make
any difference. I took a different approach and put a sniffer on the
machine being queried and see that no SNMP GET is ever happening! I can
do an snmpwalk of the same thing and decode the conversation though.
Do you have any other ideas?
Glen Warn
PEMCO Corporation Computer Services (PCCS)
glen.warn@pemcocorp.com
206-628-5770
-----Original Message-----
From: owner-nv-l@lists.us.ibm.com [mailto:owner-nv-l@lists.us.ibm.com]
On Behalf Of James Shanks
Sent: Thursday, April 28, 2005 5:31 AM
To: nv-l@lists.us.ibm.com
Subject: Re: [nv-l] Ruleset ? using Compare MIB Variable
Glen,
Have you been tracing nvcorrd's execution of your rulesets with
"nvcdebug -d all" as I suggested?
If you want to see why one case is working and the other not, that's the
way.
By definition, SNMP TimeTicks are in hundredths of a second, or at least
that's what they should be. So you should get 600 to the minute, not
1000.
There is no good source for writing rulesets. The only Redbook I know
of which discusses them is very old, SG24-4515-00, Examples Using
NetView for
AIX Version 4, which came out when rulesets were new (December 1995).
If
you search the nv-l archives you should also find a short document I
wrote giving some performance hints and tips. We've already discussed
much of it
in your last few postings. But running the trace is the way to answer
most questions.
If you are using NetView traps, then I would specify attribute 2, the
fully-qualified hostname, rather then the Origin as the source of your
MIB query. Why? You don't want to query the NetView box, but the host
the trap is about. If you are using non-NetView traps then Origin is
required.
HTH
James Shanks
Level 3 Support for Tivoli NetView for UNIX and Windows Tivoli Software
/ IBM Software Group
|