CPU and MEM use by process/app on linux

[Anon Anon]

Hi
Been thrown an interesting question

Can we report (graph and export as a csv ) stats on CPU and MEM use by process or apps on Linux hosts

We have sysedge installed so thats available

 

Any tips, pointers etc greatfully recevied

[Anon Anon]

I've done this on CA PC, but not on eHealth, also I was using the native Linux SNMP agent and not a SysEDGE agent.

 

Each process will have its own entry under the HOST-RESOURCES-MIB, hrSWRunPath is the process, and hrSWRunPerfCPU / hrSWRunPerfMem is the cpu and memory utilisation.

These can be matched up with the index

[Justin_Kulikowski]

Did you need to create a new metric family and/or vendor cert for this in CAPC, or is it already certified?

 

At what interval did you decide to run Change Detection for this metric family?  I'm imagining this SNMP table is fairly dynamic since the Index appears to be the value of the PID, which will change every time a process is started.

[Legacy User]

Looking at doing this as well.  How did you accomplish this?