Does anyone know the credentials? Or how to set this up?
Many thanks
We don't currently have SNMP polling access (or other monitors) available for the LEM appliance. What would be useful to you to monitor? Just up/down? Anything else?
Nicole,
We are looking to monitor the status of the appliance (Up / Down) Along with the other items listed below:
CPU Usage
Memory Usage
Hard Drive Usage (Most important!)
Network Usage
In addition if we could also monitor things like, Concurrent users logged into LEM, Processing Queue depth. There are probably more but that's all that comes to my head at the moment.
@nicole pauls - Is there any signs of the ability to permit SNMP polling of the LEM appliance in an upcoming release this year? It has been mentioned in other posts, but the ability to set the community string (or just set it to something other than public and state this in the user guide in line with security best practice) would be a welcome feature. Explaining to users that they cannot change the string as per recommendations is tough to swallow.
Any update on this? For example, vSphere is showing an alarm on the LEM virtual appliance for CPU usage. I need to know what the CPU usage is.
Looks like no update in LEM 6.2 (current release). The ability to poll the appliance via SNMP is possible, but this would result in countless support cases and remote sessions with SolarWinds support to login as root and configure smnpd.conf with your desired community string and permitted managers (if this service has not been removed from the OS build). If you are getting CPU alarms triggered in vSphere, check why your appliance is running hot. This can be down to number of log sources, bad syslog configuration (multiple syslog sources from different vendors to same facility level, badly formed rules or out of the box rules not tuned/disabled, no noise filtering applied.
The 2 vCPU, 8GB Memory configuration should suffice for most instances and it is key to make sure that the LEM Manager is properly configured, as per above hints. If you have a large quantity of events per minute, high node count (250+), it may be worthwhile reviewing if a 4vCPU configuration would be of benefit, but review your setup first before throwing more resource at the issue.
