We are looking at moving the majority of our workloads to the cloud and I am looking for real-world examples of how much egress bandwidth is used when we move our monitoring to agent based in the cloud. I want to be eyes wide open on the costs of monitoring and see if there are any other solutions.
I have found some examples on thwack that the agent will typically use 1.4 Kbps for a typical monitoring profile. This would equate to about 120 MB / day or 3.65 GB per month per monitored node. Does this live up to what anyone else has seen for their monitored workloads?
I have a Windows 2016 and Linux 7.4 test node that is only running the agent and I am seeing 5 - 13 Kpbs transmit.
Other ideas we are thinking of are have an APE in Azure and just sending the total traffic back to the onsite DB. Anyone have any experience with monitoring in Azure and costs?
