Oracle Communications Performance Intelligence Center (PIC) Software
 

PLA-Support>Sun Systems>CommsGBU>Global Signaling Solutions>SN-SND: Tekelec PIC
 

Applies to:

Purpose

This document describes main steps to troubleshoot problem for SNMP forwarding by NSP.

Troubleshooting Steps

Steps below must be followed to find the cause of SNMP traps not being sent by NSP:

1. Check the SNMP configuration (in platcfg menu or directly in file /etc/snmpd.conf), if not configure it.
   
2. Verify if traps are being sent from the NSP weblogic1 and weblogic2. As root type the following command on both servers (adapt ethxx to server interface):
   
   tcpdump -i ethxx -f 'port 162' -vvv
   or tcpdump -i any host <SNMP agent IP> -v
   Expected output (numbers will differ):
   tcpdump: verbose output suppressed, use -v or -vv for full protocol decode
   listening on eth01, link-type EN10MB (Ethernet), capture size 96 bytes
   0 packets captured
   1 packets received by filter
   0 packets dropped by kernel
   If any traps are being forwarded you will see them in the tcpdump output. Of course some alarms must be received by the NSP before they are forwarded.
3. Check JMS queues:
   1. Login to NSP Weblogic console http://<ip_of_nsp>/console
   2. Click Services => Messaging => JMS Servers => NSPJMSServer1a => Monitoring tab => Active Destinations tab
   3. Check each queue knowing that events are following this path:
      Messages Current/Pending must not increase and, in normal situation, is even equal to 0.
      1. JMX2ListernerQueue
      2. EventListenerQueue
      3. EvaluationObjectQueue
      4. JMSForwardingQueue
      5. jmsSNMPQueue.
         If this queue is filling, it means that weblogic is probably not able to contact the jmxd service on Weblogic1 to emit the SNMP traps . Check application.log for possible causes.
   4. Repeat steps b and c for each JMS servers:
      1. On one box NSP: NSPJMSServer1a and NSPJMSServer1b
      2. On four boxes NSP: NSPJMSServer1a, NSPJMSServer1b, NSPJMSServer2a and NSPJMSServer2b
4. Possible actions to fix the issue:
   1. Restart jmxd service on weblogic1
   2. Restart the weblogic instances to force them to reconnect to jmx (this can be done from weblogic console or simply by restarting nspservice)
      

References

This solution has no attachment