I'm trying to create error handlers to monitor partition usage so I can be alerted (email syslog or snmp) if available space drops below a specific cap. Does anyone I've accomplished this?
Incident "22" is automatically fires if any partition uses more than 90% of the available disk space. It does not allow to specify the amount (yet), but it allows to very easily generate an SNMP Event or trigger an eMail. This can be used to alert administrators to have a look. In the Event/eMail you can probably sent the Host name, so that the recipient knows which system to look at.
certainly, I attached an example XML. It is very very simple and just performs the following:
If any partition uses > 90%, it sends an eMail which contains "Partition full - please check file system".
You certainly want to apply changes to this Rule Set. Unfortunately the "90%" threshold cannot be adjusted at the moment.
I hope it gives you some idea.
AndreNachricht geändert durch asabban on 26.07.11 02:51:32 CDT
Great, looks like i was almost right thanks. Is there a way to vary the threshold? say you want an email at 80% instead of 90% which is what the incident.id is set to.
Ok thanks for that. Now i have seen how your rule works i will try and do soemthing with it. I saw an incident ID (20 i think) that reports Raid problems. Is that accruate? In the past we've had HDD failures on V6 and had no way to know about it. IF V7 can reort this problem it would be great.