The latest Network Management Pack does not include Linux and Unix systems as network devices, which is the first hint that monitoring via SNMP for these systems is not supported. An exception are following OIDs:
.126.96.36.199.4.1.9070.1.2.3.1.5 Host Linux 2610mvl401
.188.8.131.52.4.1.2312 Host Linux Linux host
.184.108.40.206.4.1.8072.3.2.10 host Linux Net-SNMP Agent 5.1
.220.127.116.11.4.1.8072.3.2.15 Host Linux NET-SNMP SOLARIS
.18.104.22.168.4.1.8072.3.2.3 host Linux NET-SNMP SOLARIS
In some cases, the Linux/unix machines are discovered as non-generic devices, but still it will be monitored as an generic device- only basic availability.
Since this is currently by design, the quickest way would be to monitor the systems with the unix/Linux SCOM agent.
If for a reason, only SNMP is allowed, custom monitors need to be created, for each OID. This is the same also for other devices we do not have in the network devices list of the NW MP.
Here is a short overview on how to configure custom monitors, in this case a monitor which reads memory usage:
Create a new monitor, at first snmpProbe, set the OID you need to query from the device:
Configure the parameter, mostly this will remain the same
You`ll get state changes, when it`s configures thresholds are met: