Zabbix and Health Checks of Server Hardware

[gglea]

How (if possible) would zabbix handle health checks of server hardware such as cpu, raid volumes, power supply etc without using SNMP traps? Is there a way for zabbix to pull these checks from the devices using the zabbix agent in order to for example show predictability of a hard disk failing? Is there a way it can work with Dell Openmanage to pull this information? Any info regarding this would be appreciated.


Thanks

[HaveDill]

For my HP servers i found a nagios vbs script that runs and reports back RAID status' using HPAcucli


So in each of my C:\Zabbix folder i put in the .vbs, then in the agentd.conf i placed the custom key i want for this script:

UserParameter = smartarray,cscript "C:\Zabbix\check_smartarray.vbs"

Then in my Zabbix host i have an item that points to that key, and checks once a day on my RAID status'. Triggers check for regular expressions in the returned value.


Part of my powershell rollout script also made shure that HPACUCLI existed on the remote machine (we are an HP house), and if not copied the proper files over so the vbs would run successfully.

[aib]

There are my five cents about RAID control:
zabbix_agentd.conf

UserParameter=raid.ad,sh /usr/lib/zabbix/externalscripts/aacraid.active_drives
UserParameter=raid.cs,sh /usr/lib/zabbix/externalscripts/aacraid.controller_status
UserParameter=raid.ld,sh /usr/lib/zabbix/externalscripts/aacraid.LD_status $1

# cat aacraid.active_drives
#!/bin/sh
sudo /usr/local/sbin/arcconf getconfig 1 PD | grep "State" | grep "Online"| wc -l

# cat aacraid.controller_status
#!/bin/sh
sudo /usr/local/sbin/arcconf getconfig 1 AD  | grep "Controller Status" | grep -vc Optimal

# cat aacraid.LD_status
#!/bin/sh
sudo /usr/local/sbin/arcconf getconfig 1 LD $1 | grep device | grep -v information

# arcconf getconfig 1 AD
Controllers found: 1
----------------------------------------------------------------------
Controller information
----------------------------------------------------------------------
   Controller Status                        : Optimal
   Channel description                      : SATA
   Controller Model                         : Adaptec 2810SA
   Controller Serial Number                 : C80489
   Physical Slot                            : 6
   Installed memory                         : 64 MB
   Copyback                                 : Disabled
   Background consistency check             : Enabled
   Automatic Failover                       : Enabled
   Stayawake period                         : Disabled
   Spinup limit internal drives             : 0
   Spinup limit external drives             : 0
   Defunct disk drive count                 : 0
   Logical devices/Failed/Degraded          : 3/0/0
   --------------------------------------------------------
   Controller Version Information
   --------------------------------------------------------
   BIOS                                     : 4.2-0 (8205)
   Firmware                                 : 4.2-0 (8205)
   Driver                                   : 1.2-0 (30300)
   Boot Flash                               : 0.0-0 (0)
   --------------------------------------------------------
   Controller Battery Information
   --------------------------------------------------------
   Status                                   : Not Installed


Command completed successfully.

[aib]

and one more - about CPU temperature:
zabbix_agentd.conf:

UserParameter=cpu.temp[*],/usr/lib/zabbix/externalscripts/cpu_t $1

[B]# cat /usr/lib/zabbix/externalscripts/cpu_t[/B]
#!/bin/bash

echo $(date '+%Y%m%d:%H%M%S') $0 $1 =\> `/usr/bin/sensors | grep temp` >> /var/log/zabbix/cpu_t.log

v=$(/usr/bin/sensors | awk -v k=$1 ' /'temp'/ {
if (k == "current")
{ split($2,a,".")
split(a[1],b,"+")
}
else
{
if (k == "high")
{ split($5,a,".")
split(a[1],b,"+")
}
}
}
END {print b[2]}')
echo $(date '+%Y%m%d:%H%M%S') $0 $1 =\> $v >> /var/log/zabbix/cpu_t.log
echo $v

[B]# /usr/bin/sensors[/B]
w83627hf-isa-0290
Adapter: ISA adapter
in0:         +1.46 V  (min =  +1.25 V, max =  +1.86 V)
in1:         +3.33 V  (min =  +2.80 V, max =  +3.79 V)
in2:         +3.31 V  (min =  +2.80 V, max =  +3.79 V)
in3:         +2.93 V  (min =  +2.53 V, max =  +3.42 V)
in4:         +3.15 V  (min =  +2.67 V, max =  +3.62 V)
in5:         +0.66 V  (min =  +0.30 V, max =  +0.99 V)
in6:         +1.78 V  (min =  +1.52 V, max =  +2.06 V)
in7:         +3.23 V  (min =  +0.54 V, max =  +3.23 V)
in8:         +1.17 V  (min =  +0.05 V, max =  +2.37 V)
fan1:       5192 RPM  (min = 1424 RPM, div = 4)
fan2:       5037 RPM  (min = 1424 RPM, div = 4)
fan3:          0 RPM  (min = 1424 RPM, div = 4)  ALARM
temp1:       +27.0°C  (high = +65.0°C, hyst = +60.0°C)  sensor = thermistor
temp2:       +28.0°C  (high = +75.0°C, hyst = +70.0°C)  sensor = thermistor
temp3:       +28.5°C  (high = +75.0°C, hyst = +70.0°C)  sensor = thermistor
cpu0_vid:   +1.500 V
beep_enable:enabled

[Raged23]

Openmange Dell Server - Windows

I use the same method as HaveDrill in a Windows enviroment.
This allows you to use an Active Agent behind firewalls.
With Open Manage installed you can use the omreport command to provide information on the Raid status.
My setup is very basic, as I don't know enough to parse the information from omreport in a more complex way.
Zabbix.agend.conf

UserParameter=raid.disks,%systemroot%\system32\windowspowershell\v1.0\powershell.exe -nologo C:\zabbix\zbx_raid_disks.ps1

zbx_raid_disks.ps1

omreport storage vdisk controller=0 | ?{$_ -match "^status"} | %{$status=1}{if($_ -notlike "*OK*"){$status=0}}{$status}

So now if your raid status is *Ok* it will return a 1. If a drive fails or any other problem it will return a 0.
You can then make a template and trigger for raid.disks.
I use this same method for intrusion,temp, etc.
I am working on a better way to parse the information to actually return values rather than just OK or Problem.
This method was taken from a much older zabbix thread.