Skip to content

Instantly share code, notes, and snippets.

@cathay4t
Created November 30, 2020 05:52
Show Gist options
  • Save cathay4t/9745acabf13430208e66c4727b00aa3d to your computer and use it in GitHub Desktop.
Save cathay4t/9745acabf13430208e66c4727b00aa3d to your computer and use it in GitHub Desktop.
Test report on VRF

Network Layout

     ns_pod1                                              ns_srv1
    veth_pod1    veth_pod1_ep            veth_srv1_ep    veth_srv1
                    |----------- vrf1 --------|
 192.168.0.1/24  192.168.0.2/24           10.0.0.2/24     10.0.0.1/24


     ns_pod2                                               ns_srv2
    veth_pod2    veth_pod2_ep             veth_srv2_ep    veth_srv2
                    |----------- vrf2 --------|
 192.168.0.1/24  192.168.0.2/24            10.0.0.2/24     10.0.0.1/24

OS Info

  • RHEL 8.3 VM
  • kernel-4.18.0-240.el8.x86_64
  • NetworkManager stopped for kernel test
  • OpenvSwitch stopped for kernel test
  • Firwalld stopped
  • Test command:
    • qperf -ip 19766 -t 60 --use_bits_per_sec 10.0.0.1 tcp_bw
    • qperf -ip 19766 -t 60 --use_bits_per_sec 10.0.0.1 tcp_lat

Kernel Test Results

With 1 pod setup like above:

qperf TCP Bandwidth:
    bw  =  48.9 Gb/sec
qperf TCP Latency:
    latency  =  12.9 us

With 1000 pods setup like above and all pingable from pod to server.

qperf TCP Bandwidth:
    bw  =  49.5 Gb/sec
qperf TCP Latency:
    latency  =  11.4 us

Nmstate/NetworkManager test

TODO

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment