How to remotely check LEDs on HDDs (Proliant servers)

[Baher Seioshansian]

Hi,

I wanted to know if there is any software or utility that enable the admins to see the status of HDDs on the server in case the servers are located in a remote site ? I need to see whether LEDs on the HDDs are RED or GREEN. I know the ACU gives alot of (text) but it's not easy to use it and time consuming .. I'm looking for a graphical package that does that. Thanks and regards,

[jspiga]

You should be able to use SNMP to give you a fairly good idea of what the LEDs on the front of each drive are saying.

OID 1.3.6.1.4.1.232.3.2.5.1.1.6 gives you the drive status:
1 = Other
2 = OK
3 = Failed
4 = Predictive Failure

OID 1.3.6.1.4.1.232.3.2.5.1.1.37 gives you the overall condition of the physical drive:
1 = Other
2 = OK
3 = Degraded
4 = Failed

OID 1.3.6.1.4.1.232.3.2.5.1.1.55 gives you the failure code (if a drive fails. 0 is returned here if there is no failure.

OID 1.3.6.1.4.1.232.3.2.5.1.1.56 gives you the amount of time (in tenths of a second) that the LEDs of each drive are blinking. If the LEDs are not blinking, 0 is returned.

OID 1.3.6.1.4.1.232.3.2.5.1.1.57 gives you the SMART status of each drive:
1 = Other
2 = OK
3 = Replace Drive

Of course, all of this returns numeric results, but it should be pretty easy to take that info and work it into a nice graphical form (It's on my low-priority list of things to do to chart these statuses in RRDTool charts but for the meantime, I have Nagios checking on each OID and notifying me if things change).

[Baher Seioshansian]

Thank you for your reply and comments.

Can you please briefly guide me on how to implement SNMP on Proliant servers ? I never used SNMP before ?

thanks and regards .. %baher%

[jspiga]

SNMP can be a very complex topic, but here goes...

If you built your server using SmartStart, SNMP should already be installed. However, you will need to configure it so that the system that you are using to monitor the server can access the SNMP info. You can do this by loading up the Systems Management Homepage on your server and allowing the monitoring host to access SNMP info from under the Tools section.

Your monitoring host will need some SNMP software installed in order to retrieve the data from your server. In the Unix/Linux realm, this generally involves installing NetSNMP and then running the snmpwalk and snmpget commands to get the SNMP data. For instance:
snmpwalk -v 1 -c public

You'll then need to work with the output to get it into your desired format.

Using the example I listed above, I use a fairly simple perl script to execute the snmpget command, process the output and return the data that nagios wants to it.

However, that's not exactly the format that you may want this data.

Creating a basic web page that shows the result should be relatively easy to write - even using PHP... But then again, I probably think that because I have been using SNMP for a few years. I do remember there being a very steep learning curve to fully understand SNMP...

[Chris Rosan]

You could also use HP Systems Insight manager. It will do this monitoring & much more for you and is reasonably easy to setup (following the instructions).