Hello all,
I am interested in using the agent instead of NET-SNMP and I've noticed a few differences regarding information, such as the agent pulling OS version for me. I've also noticed on a linux-agent monitored node I am able to see disk queue length, but on my snmp machine I can not. I believe monitoring with the agent will continue collecting data if the node loses connection with my poller? I also read where aLTeReGo said in a post the Linux agent monitoring of memory utilization is far more accurate than NET-SNMP. I think i have also read you do not need to mess with OiDs when using the linux agent? Assuming this is all true, I am interested in this. On a side note, it seems to only use a bit less than 30 megs of memory as well, has anyone observed more? I'm only testing it on one machine atm.
My big question is, is there a list of pros/cons to using linux agent vs snmp monitoring? This is something I would have to get SysAdmins on board with for it to happen. Thank you!