On the Meraki MX device under Security Appliance > Monitor > Appliance Status on the Uplink tab there is a spot showing Packet Loss on the circuit. If you hover over the peaks you are able to see the loss percentage. Do you think it would also be beneficial to show what type of packets were being lost without having to use the the Network-Wide > Monitor > Packet Capture tool? For example, in some instances and configurations TCP traffic is metered in a way that when UDP is topping the bandwidth TCP takes the hit on loss to compensate for increased UDP traffic. If the configuration is incorrect, rather than load balancing to even out the bandwidth, TCP will increase loss indefinitely to accommodate the increasing UDP traffic until it is lost completely. It would seem beneficial that when hovering over loss peaks we could identify the percentage of different packet traffic types without having to do a separate capture in order to effectively manually load balance traffic. If the loss was caused due to problems with the circuit or equipment it would all appear in similar percentages and rates as opposed to specific traffic loss. This would save time on troubleshooting and configuration errors.
@NFL0NRDefinitely, one of the big issues we end up having is with VoIP. Unless we are using a Meraki switch we get virtually no analytics out of the devices. This becomes especially frustrating when we are doing something like Telehealth in addition to normal VoIP traffic. Distinguishing between the two as far as loss goes is virtually impossible on most of our interfaces.
Please understand that the loss graph you are referring to isn't monitoring ALL packet loss, it is referencing packet loss to the IP address designated (188.8.131.52 by default) via ICMP test packets we send to the designated IPs.