Netgate Discussion Forum
    • Categories
    • Recent
    • Tags
    • Popular
    • Users
    • Search
    • Register
    • Login

    Test TNSR

    Scheduled Pinned Locked Moved TNSR
    6 Posts 3 Posters 962 Views
    Loading More Posts
    • Oldest to Newest
    • Newest to Oldest
    • Most Votes
    Reply
    • Reply as topic
    Log in to reply
    This topic has been deleted. Only users with topic management privileges can see it.
    • S
      sadekyo1712
      last edited by

      Hello everyone, im newbie with TNSR. Lastweek, i have downloaded TNSR Home+Lab and try to test performance. My setup homelab have 2 boxes. Box 1, i have installed Pfsense with 2 port 10Gbps SFP+ . Box 2, i have installed TNSR, it also have 2 port 10 Gbps SFP+. Afterthat, i config PFsense as DHCP server for 10 Gbsp port and connect it to port 10 Gbps of TNSR box. I also config firewall rule in pfsense to allow all traffic. With TNSR, i config dataplane to accept 10 Gbps interface (successfully get ip from dhcp server) and add ACL to permit all traffic (You can see in the attachment). But when i try to test performance of TNSR by using iperf3 (PFsense as a client and TNSR as a server), the bandwidth for TCP test is just arround 3.47Gbps. I dont know what wrong or how tuning to improve performance, because i see the result is too low for my requirement (about 8 or 9 Gbps if no firewall rule is applied)

      • TNSR acl
        Screen Shot 2020-09-22 at 09.06.12.png

      • TNSR 10 Gbps interface
        Screen Shot 2020-09-22 at 09.06.29.png

      • PFsense 10 Gbps interface
        Screen Shot 2020-09-22 at 09.07.05.png

      Thanks in advance!

      1 Reply Last reply Reply Quote 0
      • DerelictD
        Derelict LAYER 8 Netgate
        last edited by

        Maybe pfSense is the slow link there. If you are going to test performance you need to be sure that the test suite is not slower than the device you are testing.

        And, as usual, when you are testing a router you should test through it not to/from it.

        Also, please explain how you ensured iperf3 was running on the tnsr dataplane and not on its host interface.

        You also didn't specify the capabilities of the hardware you were using, such as CPU.

        Chattanooga, Tennessee, USA
        A comprehensive network diagram is worth 10,000 words and 15 conference calls.
        DO NOT set a source address/port in a port forward or firewall rule unless you KNOW you need it!
        Do Not Chat For Help! NO_WAN_EGRESS(TM)

        1 Reply Last reply Reply Quote 1
        • S
          sadekyo1712
          last edited by

          @Derelict i use 2 boxes that are identical and have the same 10Gbps sfp+ module so the test suite is not slower than the device i am testing.

          I know use 2 boxes to test is not the perfect way to test bandwidth but in my homelab condition, i dont have optical switch and i just have only 2 10Gbps sfp+ module to test Gigabits network performance, so it is the only way for me .

          • "Also, please explain how you ensured iperf3 was running on the tnsr dataplane and not on its host interface."
            => i install iperf3 in TNSR device. In tnsr cli, i use the command "dataplane shell" to go shell, next i can see my NIC with ip address with ifconfig command
            Screenshot_20200928_155827.png
            and then i run iperf3 server from here

          The hardware of my device is :

          • Intel(R) Atom(TM) CPU C3558 @ 2.20GHz 4 CPUs

          • Ram 8 GB

          • SSD 256Gb

          1 Reply Last reply Reply Quote 0
          • DerelictD
            Derelict LAYER 8 Netgate
            last edited by

            i use 2 boxes that are identical and have the same 10Gbps sfp+ module so the test suite is not slower than the device i am testing.

            The one running pfSense still cannot necessarily pass 10Gbit/sec. That's why tnsr exists in the first place.

            Why not put tnsr on both?

            Unless you can pass traffic through tnsr instead the test still feels invalid to me since you are using it as a traffic processor (generator/receiver) and a router at the same time.

            Chattanooga, Tennessee, USA
            A comprehensive network diagram is worth 10,000 words and 15 conference calls.
            DO NOT set a source address/port in a port forward or firewall rule unless you KNOW you need it!
            Do Not Chat For Help! NO_WAN_EGRESS(TM)

            S 1 Reply Last reply Reply Quote 0
            • S
              sadekyo1712 @Derelict
              last edited by

              @Derelict said in Test TNSR:

              The one running pfSense still cannot necessarily pass 10Gbit/sec. That's why tnsr exists in the first place.

              Why not put tnsr on both?

              Unless you can pass traffic through tnsr instead the test still feels invalid to me since you are using it as a traffic processor (generator/receiver) and a router at the same time.

              Yes i know, right now i am setting up new lab to test bandwidth through tnsr, i will re-inform in this thread when i have finished my test lab. Thanks you so much.

              audianA 1 Reply Last reply Reply Quote 0
              • audianA
                audian @sadekyo1712
                last edited by

                @sadekyo1712 - Thanks, look forward to your updates

                1 Reply Last reply Reply Quote 0
                • First post
                  Last post
                Copyright 2025 Rubicon Communications LLC (Netgate). All rights reserved.