Thread: Nagios doesn't get plugin's output

Hello,

We have installed the mtk-nagios-1.1 plug-in from rpm and also the sources, which contain the sample files.

We have successfully restarted Nagios with all the services configured for one host, but when seeing the Web GUI, the services show (No Output), and the status becomes WARNING.

This is the output for the fan-data option, when run manually:

/usr/lib64/nagios/plugins/check_dell_status -H w.x.y.z --fan-data
ESM MB Fan1: 4800 RPM (OK) ESM MB Fan2: 4920 RPM (OK) ESM MB Fan3: 7320 RPM (OK) ESM MB Fan4: 7140 RPM (OK) ESM MB Fan5: 7320 RPM (OK) ESM MB Fan6: 7200 RPM (OK) ESM MB Fan7: 7320 RPM (OK)

In the web GUI, this is what we get:

Dell Server Fans
WARNING
06-24-2009 18:32:32
0d 0h 42m 18s
5/5
(No output!)

Your advice or comments will be greatly appreciated.

The plug-in apparently works very good except for this issue. It's been also posted in Dell's site:
Dell's Nagios Plugin

Thank you and regards,
Max Glucksmann
Comtel Networks, Corp.
maxglucksmann@hotmail.com

Hello,

Just wanted to make an observation...

After setting the SETUID bit of the other files contained within the RPM, output was generated.

Still, for the storage, virtual disks and battery items, the message "NO OMSA INSTALLED" was received.


chown root.root /usr/lib64/nagios/plugins/dell_health.*
chown root.root /usr/lib64/nagios/plugins/check_dell_status

chmod 755 /usr/lib64/nagios/plugins/dell_health.*
chmod +s /usr/lib64/nagios/plugins/dell_health.*

chmod 755 /usr/lib64/nagios/plugins/check_dell_status
chmod +s /usr/lib64/nagios/plugins/check_dell_status

ls -l /usr/lib64/nagios/plugins/*dell*
-rwsr-sr-x 1 root root 4211 2009-06-24 22:01 /usr/lib64/nagios/plugins/check_dell_status
-rwsr-sr-x 1 root root 53780 2009-03-17 15:34 /usr/lib64/nagios/plugins/dell_health.py
-rwsr-sr-x 1 root root 39594 2009-03-17 15:34 /usr/lib64/nagios/plugins/dell_health.pyc
-rwsr-sr-x 1 root root 39594 2009-03-17 15:34 /usr/lib64/nagios/plugins/dell_health.pyo



After checking the wrapper script "check_dell_status", to avoid the "NO OMSA INSTALLED" message, I created a service calling directly the "dell_health.py" script.

Now, I'm getting "No value available" messages for the storage arrays, virtual disks and batteries...

This is the service:

nano /etc/nagios/commands/check_dell_health.cfg

# 'check_dell_health' command definition
define command{
command_name check_dell_health
command_line $USER1$/dell_health.py --host=$HOSTNAME$ --$ARG1$
}


Just to confirm, this is the output of the disk array, which doesn't show in the GUI:

/usr/lib64/nagios/plugins/check_dell_status -H w.x.y.z --storage-array-data
SCSI: Physical Disk 0:0: [Available] (OK) SCSI: Physical Disk 0:1: [Available] (OK) SCSI: Physical Disk 0:2: [Available] (OK) SCSI: Physical Disk 0:3: [Available] (OK) SCSI: Physical Disk 0:4: [Available] (OK)



Or using directly the dell_health.py:

/usr/lib64/nagios/plugins/dell_health.py --hostname=w.x.y.z --storage-array-data
SCSI: Physical Disk 0:0: [Available] (OK) SCSI: Physical Disk 0:1: [Available] (OK) SCSI: Physical Disk 0:2: [Available] (OK) SCSI: Physical Disk 0:3: [Available] (OK) SCSI: Physical Disk 0:4: [Available] (OK)


Any lead will be greatly appreciated.

Thank you very much,
Max

Hello,

Just a quick update...

We were not using real hostnames, so by changing $HOSTNAME$ to $HOSTADDRESS$, the values showed properly.

This one is a real bug though, in the section that shows the voltage-data:

/usr/lib64/nagios/plugins/check_dell_status -H w.x.y.z --voltage-data
Traceback (most recent call last):
File "/usr/lib64/nagios/plugins/dell_health.py", line 1311, in <module>
app.parse()
File "/usr/lib64/nagios/plugins/dell_health.py", line 278, in parse
str, result = self.getVoltageInformation()
File "/usr/lib64/nagios/plugins/dell_health.py", line 700, in getVoltageInformation
str += "%.1f V" % (voltages[x]['Reading'] / 1000)
TypeError: unsupported operand type(s) for /: 'str' and 'int'


BR,
Max

/usr/lib64/nagios/plugins/dell_health.py --hostname fileserver-01 --voltage-data
Traceback (most recent call last):
File "/usr/lib64/nagios/plugins/dell_health.py", line 1311, in ?
app.parse()
File "/usr/lib64/nagios/plugins/dell_health.py", line 278, in parse
str, result = self.getVoltageInformation()
File "/usr/lib64/nagios/plugins/dell_health.py", line 700, in getVoltageInformation
str += "%.1f V" % (voltages[x]['Reading'] / 1000)
TypeError: unsupported operand type(s) for /: 'str' and 'int'


On another system however it does work, and the only difference I can think of is that the one that it works on has a DRAC card, and the one that it fails for does not.

The bug above appers on only certain systems, I have seen in on R900 and PowerEdge 2950 systems. Older PowerEdge 2950 does not seem to be affected, but I am not 100% sure it is not that the older PowerEgde 2950 I have tested with has a DRAC card.

The problem is that the code tries to divide a string. The below patch solved the problem.

Code:

--- dell_health.py	2009-07-20 20:00:36.000000000 +0100
+++ dell_health.py.test	2009-07-20 20:04:21.000000000 +0100
@@ -697,7 +697,8 @@
             if voltages[x]['Reading'] == []:
                pass
             else:
-               str += "%.1f V" % (voltages[x]['Reading'] / 1000)
+	       voltages[x]['Reading'] = voltages[x]['Reading'].strip('mV')
+               str += "%.1f V" % (int(voltages[x]['Reading']) / 1000)
             voltages[x]['Voltage Status'], stateFlag = self.getStatusProbe(voltages[x]['Voltage Status'])
             if stateFlag >= prevState:
                prevState = stateFlag