How do you monitor critical system error domain wide ?

[albertwt]

Greetings everyone,

 

How do you organize or implement monitoring system in your domain to monitor these critical error on your workstations and servers ?

 

    Hardware Related Error:
   7, 9, 11, 51, 52, 55 Potential HD related issue 
   1053 Servers too hot. Sometimes our Air conditioning breaks.

 

Any kind of help and comments will be greatly apprecaited.

 

Thanks,

 

AWT

[Ric_]

You need something to catch the errors and notify you. The most widely used systems are Nagios and Systems Center Operations Manager (SCOM). Both take some setting up mind.

[albertwt]

Nagios :-o

wow seriously can Nagios do that ? i thought only checking the status of the network eg. ping at certain interval.

 

Thanks for the reply.

[chrisbrown]

Nagios :-o

wow seriously can Nagios do that ? i thought only checking the status of the network eg. ping at certain interval.

 

Thanks for the reply.

nagios can do EVERYTHING. You can even attach it to temperature sensors and integrate with mrtg. This will give you graphs of your server room temperature and alerts when over a specified threshold. Easy as to configure too.

[jsnetman]

Server 2008 and win 7 clients you can configure event forwarding and event subscriptions. Basically this forwards any type of event to a central machine. But other than that I agree SCOM. Tried nagios and could not get to grips with it, but saying that I did not really try that hard to get it configured correctly.

[bio]

cacti with nagios can do the trick

 

bio..

[pete]

nagios can do EVERYTHING. You can even attach it to temperature sensors and integrate with mrtg. This will give you graphs of your server room temperature and alerts when over a specified threshold. Easy as to configure too.

 

Yup. If you can get snmp or any form of text/numerical data out of something you can monitor it in Nagios. It might occasionally take several goats and you'll have to re-flake the flint blade a couple of time, but it's usually doable.

[albertwt]

ok, thanks guys for your reply.