Net-snmp issue in Solaris 8

Hi there,

I know this is a common question, but I have not found a satsifactory solution yet, so pardon me for posting similar questions.

My problem is the log /var/adm/messages is filling up with this entry every few minutes Please refer below.
Oct 4 21:39:46 host last message repeated 3 times
Oct 4 21:41:47 host /usr/lib/snmp/snmpdx: [ID 702911 daemon.error] bind() failed on UDP port 161 [errno: Address already in use(125)]

System info :SunOS xxxxxx 5.8 Generic_117350-44 sun4u sparc SUNW,Sun-Fire-V240

Now I have already disabled the native Sun's snmp daemons, and using Net-snmp

edited list

/etc/rc0.d: moved_K07dmi moved_K07snmpdx
/etc/rc1.d: moved_K07dmi moved_K07snmpdx
/etc/rc2.d: moved_K07dmi moved_K07snmpdx
/etc/rc3.d: moved_K07dmi moved_K07snmpdx
/etc/rcS.d: moved_K07dmi moved_K07snmpdx

Net-snmp starts using /etc/init.d/netsnmp script which is as follows

case "$1" in
'start')
/usr/local/sbin/snmpd
echo "Daemon snmpd started"
;;
'stop')
/usr/bin/pkill -9 -x -u 0 '(snmpd|snmpv2d|mibiisa)'
echo "Daemon snmpd stopped"
;;

When I do lsof I get the following result
snmpd 2264 root 8u IPv4 0x3000dbe1e70 0t0 TCP *:199 (LISTEN)
snmpd 2264 root 9u IPv4 0x3000573edb0 0t0 UDP *:62775 (Idle)
snmpd 2264 root 11u IPv4 0x3000dbe0c70 0t0 UDP *:161 (Idle)

ps return

ps -ef| grep snm
root 2264 1 0 21:00:01 ? 0:03 /usr/local/sbin/snmpd

When I do snmpget I receive this

[host] $ snmpget -v 1 -c xxxxxxx localhost sysUpTime.0
DISMAN-EVENT-MIB::sysUpTimeInstance = Timeticks: (344667) 0:57:26.67

snmpwalk returns
snmpwalk -v 2c -c xxxxxxx -m ALL localhost .1.3 | more
SNMPv2-MIB::sysDescr.0 = STRING: Sun SNMP Agent, Sun-Fire-V240
SNMPv2-MIB::sysObjectID.0 = OID: NET-SNMP-TC::solaris
DISMAN-EVENT-MIB::sysUpTimeInstance = Timeticks: (352998) 0:58:49.98
SNMPv2-MIB::sysContact.0 = STRING: "System Admin"
SNMPv2-MIB::sysName.0 = STRING: xxxxxx
SNMPv2-MIB::sysLocation.0 = STRING: LOT28
SNMPv2-MIB::sysServices.0 = INTEGER: 0
....

snmpd -f -Le returns

[host] $ snmpd -f -Le
[init_smux] bind failed: Address already in use
Error opening specified endpoint ""
Server Exiting with code 1

Why does the log still records the same errors.
And why does net-snmp's snmpd -f -Le says bind failed?

Please assist. I do not know much about Net-snmp or snmp for the matter.

The K stands for "kill". You have disabled Sun's scripts to shutdown Sun's snmp. The S links are for startup. Try disabling them too. What is your motive for switching to a new snmp server?

thanks for the quick reply. I will check it out.

About the motive, there is none. Just wanted Solarwinds to monitor the Sun machines.
Just didn't know how to setup Sun snmp. So am using net snmp.

Btw, how to setup Sun's snmp? :slight_smile:

S link script for sun's snmp was already renamed (disabled).
rc3.d:
s76snmpdx s77dmi <<<Sun's Snmp startup script link files
S99netsnmp <<<netsnmp startup script link

Hemm,, seems no is able to assist..

Well, the error messages are gone now. :slight_smile:

I renamed mibiisa.* files and restarted snmpd and it worked. No more annoying messages in the log.