Hitachi

JP1 Version 12 JP1/Extensible SNMP Agent Description, Operator's Guide and Reference


7.4.3 Acquiring detailed trace information about the master agent

The detailed trace information about the master agent includes the following information:

You can use the detailed trace information output for the master agent to determine the cause of a communication failure between the master agent and the manager or between the master agent and a subagent. The following procedure shows how to acquire detailed trace information about the master agent.

Procedure

  1. Stop snmpdm.

    Execute the following command to send a termination signal to the snmpdm process:

    Kill -15 snmpdm-process-ID

    Use the following method to obtain the snmpdm process ID.

    Execute the following command:

    /opt/CM2/ESA/bin/snmpcheck

    Execution of this command displays a list of SNMP Agent processes similar to the following. The value listed for pid= in the snmpdm line is its process ID.

    snmpdm  running pid=11293
    mib2agt running pid=11330
    hp_unixagt      running pid=11331
    ipv6agt         running pid=11334
    trapdestagt     running pid=11348
    extsubagt       running pid=11384
    htc_unixagt1    running pid=11366
    htc_unixagt2    running pid=11367
    htc_unixagt4    running pid=11368
  2. Start the master agent.

    The logs can be output to any file. For details about the log output destination, see 7.3.2 Log output destination in 7.3 Collecting logs.

    SR_SNMP_TEST_PORT=port-number-of-SNMP-Agent (typically, 161)
    export SR_SNMP_TEST_PORT
    /usr/sbin/snmpdm -tcplocal -hexdump -vbdump -apall -n >> /tmp/esa.packet.log 2>&1 &
  3. Start mib2agt and hp_unixagt (for HP-UX (IPF)).

    /sbin/init.d/SnmpMib2 start
    /sbin/init.d/SnmpIpv6 start
    /sbin/init.d/SnmpHpunix start
  4. Verify that snmpdm is running.

    /opt/CM2/ESA/bin/snmpcheck
    Important

    /tmp/esa.packet.log is a file that is continually added to. If you are monitoring file sizes and the size of this file grows too large, you can delete it as follows (for Solaris, run in ksh):

    cat /dev/null > /tmp/esa.packet.log

For Solaris and AIX, if the user logs out of the terminal on which this command was executed while detailed trace information is being acquired, a HUP signal is sent to the process in which this command was generated, which might prevent detailed trace information from being acquired. This HUP event does not occur if you do not log out of the terminal. If you must log out of the terminal, use the following method:

After you execute the detailed trace information command, find its process ID, and then execute the following command:

 /usr/bin/nohup -p -a "PID-of-process"

For details about the nohup command, see the man pages.