Forum Discussion

Slingshotz's avatar
8 years ago

Collector Bandwidth Utilization

Can anyone here tell me approximately what amount of data usage they see between collectors and the hosts?  We have a collector setup in our main office monitoring about 50 devices (mostly Dell servers and Hyper-V VMs) and these devices are all over Canada connected through our Meraki VPN mesh network.  I can see from the Meraki utilization reports that each device consumes between 1 to 2 GB of data PER DAY over port 49154 while it is reporting back to the collector.  This seems excessively high as their documentation says the average is 3.2Kbps per device which equals about 35 MB per day.  I do understand that physical hardware devices have more data to report than a Microsoft VM but even a plain Microsoft VM still consumes about 500MB/day which still seems really high.  

I have looked at our Collector settings and I don't really see any settings that might cause this and we are mostly running with the default settings when we installed LogicMonitor as it worked pretty well from the default.  Most of the servers have little to no errors so it is not spamming the collector with warnings and issues either.

  • The normal LogicMonitor deployment model assumes that the Collector is local to the devices it is monitoring.  No question, the data transfer between the Collector and the devices can be somewhat heavy, especially for protocols such as WMI (which is how LogicMonitor talks to Hyper-V).  Once the data is at the Collector, it is reduced to numerical format and compressed before forwarding to the LogicMonitor Datacenter.  The compressed Collector-to-Datacenter stream should be much smaller than the original Collector-to-Target stream.  

    If you have the ability to look at Netflow data on your primary gateway, compare the utilization from Collector to Datacenter to what you are seeing from the Meraki utilization report.  I suspect it will be much closer to the documented numbers.