[Nagiosplug-help] Snmp check disk question

[Subhendu Ghosh]

On Wed, 5 Mar 2003, Pietro Bandera wrote:

> 
> Hi all
> 
> I'm using snmp in order to take a look on the many devices/software i
> have on my lan/wan and i need to know a thing:
> 
> To control if a disk is broken or not i have found two ways...one is
> perfect the other one ...i need help
> 
> 1) the perfect one:
> 
> On the compaq server i installed the compaq agents and then ask them the
> "state" of the disk using the correct snmp OID 
> 
> 2) the unsure one:
> 
> On all the others server (IBM, SUN, BULL, etc...) i have, i can only
> control the disk using the standard HOST-RESOURCES-MIB and by the OID
> host.hrDevice.hrDeviceTable.hrDeviceEntry.hrDeviceStatus."number of the
> device i want to check"
> 
> But by this OID i can't check the disk itself i can only check the array
> device!...so
> 
> What i need to know is: if a disk of the array fails, does the
> host.hrDevice.hrDeviceTable.hrDeviceEntry.hrDeviceStatus."number of the
> array device" become warning(3) or not???????
> 
> If it doesn't do you know a better way to check the status of a disk!
> (fail or run)
> 
> Thanks all
> 
> Pietro

HOST-RESOURCES can only monitor "mounted" devices so it will only see the 
array device not the array components. Unless the array components have 
there own snmp mib, you will have to create your own scripts.

The easiest way is to use the net-snmp agent and extend it by adding 
commands that output the status of disks in an array.

See the EXTENSIBLE-MIB section in snmpd.conf


-- 
-sg