Shut down the agents and SNMP daemon:
Enter /etc/init.d/hpima stop, and then press Enter.
Enter /etc/init.d/snmpd stop, and then press Enter.
Enter ping localhost, and then press Enter to test connectivity to the local host.
If the command fails, there is a problem with your network configuration, which you must investigate and repair before continuing with agent troubleshooting.
For each <remote host name> in the snmpd.conf file (described in “First Troubleshooting Steps”), enter ping <remote host name>, and then press Enter to test connectivity to the remote management station.
 |
| |
|
 | NOTE: If your system is running RHEL, this file is located in /etc/snmp/snmpd.conf. If your system is running SLES, this file is located in /etc/snmpd.conf. |
|
| |
|
If the command fails, there is a problem with your core network configuration or your network in general, which you must investigate and repair before continuing with agent troubleshooting.
Open a new window, enter snmpd -f -L -Cc /opt/hp/hpima/etc/snmpd.conf.open, and then press Enter. This starts the SNMP daemon that uses a simple configuration file that comes with HP Insight Management Agents.
The configuration file in the command uses the read-only community string “public” for any host on your network. It is insecure but useful for troubleshooting.
Ensure that you see the following output, which verifies that SNMP has AgentX functionality. Without this functionality, the HP Insight Management Agents cannot communicate through SNMP.
Turning on AgentX master support.
NET-SNMP version <x.x.x>
|
Wait at least ten seconds to ensure the daemon continues to run. Be aware that the next step might cause output to appear in this window. Examine the output for errors if the operations in the next step fail.
On the system that has the HP Insight Management Agents installed, enter:
snmpget -Oqs -v1 -c public localhost sysDescr.0
where:
Then press Enter to test local SNMP functionality.
|
| |
|
| NOTE: The first option in the command is the capital letter “O” and the final argument ends with a period followed by a zero. |
|
| |
|
The output from this command should be a one-line response.
Enter uname -a, and then press Enter.
Compare the output of uname -a with the output you noted. The responses should largely match. If they do not, you have a malfunctioning SNMP package. Remove the HP Insight Management Agents and reinstall.
Check to see if your network is passing SNMP packets.
Log in on any of the remote hosts in your snmpd.conf file and execute a simple SNMP GET command.
If the GET command fails, especially if there is no output in the logging window, your network is not passing SNMP packets. Repair your networking issues before continuing with agent troubleshooting.
The HP Insight Management Agents depend on a running SNMP daemon. To check and troubleshoot the daemon status on the managed host, restart the SNMP daemon with the real configuration file:
Enter /etc/init.d/snmpd restart, and then press Enter.
The logging window should indicate that the previous daemon stopped. Close that window and wait at least ten seconds.
Enter /etc/init.d/snmpd status, and then press Enter to check SNMP daemon status.
You should see output similar to the following:
snmpd (pid xxxxx) is running... |
If snmpd is not running, see the snmpd(8) manpage for information on how to turn on “logging to a file” and follow the instructions to start the SNMP daemon.
If, while running RHEL, the values for PCI I/O slot numbers appear to be invalid (for example, PCI devices are not mapped to the physical slots in which they are installed), check the /var/log/hp/hpima file for the following message:
Warning: hpima version <x.x> does not support multi-domain PCI
|
If you find this message, you must reset your ACPI configuration value to single-pci-domain.
See the Red Hat Enterprise Linux Support Notes for further information about this problem and instructions on how to reset the ACPI configuration value. You should have received this document with your RHEL purchase from HP. It is also available at the following Web site:
www.docs.hp.com/en/linuxredhat.
Remove the agent log file /var/log/hp/hpima (if it exists) to clear old messages.
Enter /etc/init.d/hpima start to start the agents.
Wait at least one minute, then enter /etc/init.d/hpima status, and then press Enter to check agent status.
Verify that you see the following output:
hpimaX (pid 29032) is running...
hpimahostd (pid 29547) is running...
hpimathreshd (pid 29557) is running...
hpimahealthd (pid 29570) is running...
hpimastdeqd (pid 29580) is running...
hpimaeventd (pid 29590) is running...
hpimaidad (pid 29608) is running...
hpimasasd (pid 29619) is running...
hpimascsid (pid 29631) is running...
hpimafcad (pid 29643) is running...
hpimanicd (pid 29661) is running...
hpimaE0traps (pid 29675) is running...
|
If any of the programs listed in the previous example are not running, examine the /var/log/hp/hpima file for errors.
Enter snmpget -Oqs -v1 -c ROlocal localhost sysDescr.0, and then press Enter to test basic SNMP functionality with the real community strings.
|
| |
|
| NOTE: The first option in the command is the capital letter “O” and the final argument is a period followed by a zero. |
|
| |
|
The output from this command should be a one-line response.
Enter uname -a, and then press Enter.
Compare the output of uname -a with the output you noted. The responses should largely match. If they do not match, your community string is not properly configured. See “First Troubleshooting Steps” for troubleshooting information.
Repeat the preceding step using the RWlocal community string.
Enter snmpget -Oqs -v1 -c ROlocal -mALL localhost cpqHoMibRevMajor.0 to retrieve an agent data value from the managed system.
You should see the following output:
Repeat the previous step with the RWlocal community string.
Printable version
