I have 3 nodes deployed which are connected via Ethernet and network goes from node0-node1-node2.
using netem, ran the simulation manually with two network profiles: 100mbit10ms | 200mbit20ms
node0: server (10.10.1.1)
node1: bridge (10.10.1.2)
node2: client/sender (10.10.2.2)
network profile is set only on bridge node and when i ping from sender to server, I get accurate time delays of 10.x ms
and 20.x ms on ping and same results when i send request via iperf3 from sender to server , while I don't change the network profile on sender at all.
But I get variation when I run ping and iperf on sender side consecutively, the bandwidth rate and delay value fluctuates and is not the same. I also set limit on netem e.g limit 42 for 100mbit rate and 10ms delay but still I am not getting accurate results.
if someone can explain why I am getting this variation and how can I eliminate in netem ?