Problems resolving OIDs with snmptrap.pl

[grenzlaeufer]

Hallo!

I am working far to long on this problem. No searching or reading manuals helped me.

I want to receive with snmptrad snmp-traps and send them by snmptrap.pl to zabbix. This works fine, but I don´t get the OIDs resolved, so it is not possible to understand the snmp-mails with a lot of OIDs.
If I send a test-trap from an ILO-Board I get this message in my snmptrapd.log:

2010-12-20 20:32:56 server-ilo.domain-so.net [192.168.1.2] (via UDP: [151.189.37.58]:1267->[151.189.37.13]) TRAP, SNMP v1, community COMPAQ
	.1.3.6.1.4.1.232 Enterprise Specific Trap (11003) Uptime: 34 days, 6:15:46.30
	.1.3.6.1.2.1.1.5.0 = STRING: host is unnamed	.1.3.6.1.4.1.232.11.2.11.1 = INTEGER: 0	.1.3.6.1.4.1.232.11.2.8.1 = STRING: "Remote Insight Test Trap"

This information is transfered to the traphandler snmptrap.pl, which posts to its debug.log:

2011-01-03 13:00:14 -- 32707 --/usr/sbin/zabbix-sender --zabbix-server 127.0.0.1 --port 10051 --host snmptraps --key snmptraps --value "(zabbix.domain.net, 151.189.37.58) , .1.3.6.1.2.1.1.3.0: 47:22:43:17.03, .1.3.6.1.6.3.1.1.4.1.0: .1.3.6.1.4.1.232.0.11003, .1.3.6.1.2.1.1.5.0: host is unnamed, .1.3.6.1.4.1.232.11.2.11.1: 0, .1.3.6.1.4.1.232.11.2.8.1: \"Remote Insight Test Trap\", .1.3.6.1.6.3.18.1.3.0: 151.189.37.58, .1.3.6.1.6.3.18.1.4.0: \"COMPAQ\", .1.3.6.1.6.3.1.1.4.3.0: .1.3.6.1.4.1.232"

If I try to resolve an OID with snmptranslate then it works:

server:/etc/snmp # snmptranslate 1.3.6.1.2.1.1.5.0
SNMPv2-MIB::sysName.0

I am somehow irritated if I start the snmpd manually:

# snmptrapd -f -Le -Dparse-mibs 2>&1 | grep SNMPv2-MIB

parse-mibs:   Module 92 SNMPv2-MIB is in /root/.snmp/mibs/snmpv2-mib.mib
parse-mibs:   Module SNMPv2-MIB already noted
parse-mibs:     SNMPv2-MIB is now in /usr/share/snmp/mibs/SNMPv2-MIB.txt
parse-mibs: Parsing MIB: 92 SNMPv2-MIB
parse-mibs: Processing IMPORTS for module 92 SNMPv2-MIB
parse-mibs: Module SNMPv2-MIB already loaded
parse-mibs: Module SNMPv2-MIB already loaded
Cannot find module (CPQOS-MIB +CPQDCEO-MIB +RFC1213 +SNMPv2-SMI +SNMPv2-MIB): At line 0 in (none)

Despite that the snmpd tries to load some modules twice, it seams to load the SNMPv2-MIB and tells later that it can´t find the module. Could this be the reason for my problem?

At last I show you my configs:

this is my snmp.conf:

server:/etc/snmp # cat snmp.conf 
mibs SNMPv2-MIB
showMibErrors 0
mibAllowUnderline 1

and my snmptrapd.conf:

traphandle default /usr/bin/perl /etc/zabbix/bin/snmptrap.pl
disableAuthorization yes
donotlogtraps  no
doNotRetainNotificationLogs yes
ignoreAuthFailure yes

my snmpd.conf:

syslocation Server Room
syscontact Sysadmin (root@localhost)
rocommunity public 127.0.0.1

This is the configuration-part of my snmptrapd.pl

use strict;
use DBI;

my $DEBUG = 1;
my $DEBUGFILE = "/tmp/zabbix_snmptrapdebug.log";
my $ZABBIX_SERVER = "127.0.0.1";                     # Hostname/IP-Address of zabbix-server
my $ZABBIX_PORT = 10051;                                # Port of zabbix-server (Default: 10051)
my $ZABBIX_ITEM = "snmptraps";                          # Default item name to send traps to
my $ZABBIX_SENDER = "/usr/sbin/zabbix-sender";    # Path to your zabbix-sender
my $WILDCARD_HOST =  "snmptraps";                       # Hostname or IP-Adress of wildcard-host within zabbix
my $SEND_ALL_TO_WILDCARD = 1;                       # Send all traps to wildcard host
my $ZABBIX_DB_HOST="127.0.0.1";
my $ZABBIX_DB_NAME="zabbix";
my $ZABBIX_DB_USER="zabbix";
my $ZABBIX_DB_PASS="SecrEtPasS";
my $TRAPEXCLUDEFILE="/etc/zabbix/traps_to_exclude";
my $ALIASFILE="/etc/zabbix/aliasfile";
my $TRIMLONGVALUES=1;
my $KEYOIDNAME="SNMPv2-MIB::snmpTrapOID.0";

Systen configuration
Linux: SLES 11 SP1
Software:
snmp-mibs-5.4.2.1-8.5.1
snmptt-1.2-1.1
libsnmp15-5.4.2.1-8.5.1
net-snmp-5.4.2.1-8.5.1

I would be very grateful if someone coult point me into the right direction!

[Ori0n]

Is your snmptrapd somehow defaulting to "-On" or something similar?

Does that error regarding the SNMPv2-MIB mib go away if you restrict the MIB directory to either /root/.snmp/mibs or /usr/share/snmp/mibs with "-M"?

[misch42]

Hi,

did you enable the additional enterprise MIBs in your setup? See
-m and -M options in "man snmpcmd" and
corresponding ooptions in "man snmp.conf"

Michael Schwartzkopff

[grenzlaeufer]

Thank you for the replies! I found the problem thanks to Ori0n. It was the "-On"-Option. I shouldn´t rely on the configuration other people built up. With snmptt everything was fine as far as I was told, but now I have to built up zabbix with snmp.
Thank you very much!

[Ori0n]

Thank you for the replies! I found the problem thanks to Ori0n. It was the "-On"-Option. I shouldn´t rely on the configuration other people built up. With snmptt everything was fine as far as I was told, but now I have to built up zabbix with snmp.

Ha, using SNMPTT is what caused me to think of the "-On" option possibly fouling things up.

On a possibly related note it is possible to integrate SNMPTT and Zabbix:

http://ori0nx.wordpress.com/2010/12/30/a-better-zabbix-trap-handler-using-snmptt/

Glad you got it worked out.

[grenzlaeufer]

Thanks for the link Ori0n, its bookmarked, but I wan´t first to take the infrastructure we have.
I have only one single question: How can I manually change the trigger status: problem to ok? sometimes i get snmp-traps about java exceptions which don´t recover in a way zabbix excepts it.

[Ori0n]

How can I manually change the trigger status: problem to ok?

You can't change it manually, but you can do it automatically. Look here for an example of how to do this with a trigger (use the nodata function):

http://www.zabbix.com/forum/showpost.php?p=77924&postcount=2