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

    Low throughput under vmware wkstn 12

    Scheduled Pinned Locked Moved Virtualization
    8 Posts 3 Posters 1.4k 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.
    • J
      JamesVA
      last edited by

      Greetings!  Initially I was struggling getting a 1gbs throughput on my pfSense build (Atom D2550, 4gb RAM, 2x1gbps NICs on a Jetway board)..  was getting about 500gbps up/down using an internet speedtest.  All services other than minimal were turned off.  CPU was pegged all the time during the test.

      So i decided to ee how much more horsepower I will need and did a fresh install on a VMware workstation 12 VM (2 vCPU (i5-3427U), 2gb RAM, 2 virtual NICs).  NICs are on the same switch, but with are sitting on different subnets.

      192.168.1.0/24 - is being used as a WAN subnet
      10.0.100.0/24 - is being used as a LAN subnet

      Ran an iPerf test from a host on 10.0.100.x subnet to a host on 192.168.1.x and was getting about 160mbps, while CPU was at about 100% utilization.

      So I added 2 more vCPUs.  At 4 vCPUs, when I ran the same test, I was still getting identical throughput (160mbps), but CPUs were only being utilized about 45%.  mbuff seemed OK.

      I've attached output of 'top -SH'.  On a side note - I did not manually install vmware tools, assuming it's already there?  Or is everyone running it on ESXi installs tools manually?

      Any advice appreciated.

      Thanks!
      2017-02-11_15-09-36.jpg
      2017-02-11_15-09-36.jpg_thumb

      1 Reply Last reply Reply Quote 0
      • J
        JamesVA
        last edited by

        Anyone have any idea?

        1 Reply Last reply Reply Quote 0
        • KOMK
          KOM
          last edited by

          What vNIC were you using, the E1000 or the vmxnet3?

          1 Reply Last reply Reply Quote 0
          • J
            JamesVA
            last edited by

            both e1000 (configured in the .vmx)

            1 Reply Last reply Reply Quote 0
            • KOMK
              KOM
              last edited by

              The e1000 NICs use more CPU if I recall.  Try with vmxnet3 instead and see if it makes any difference.

              1 Reply Last reply Reply Quote 0
              • J
                JamesVA
                last edited by

                Switched to vmxnet3, but utilization and throughput are identical to e1000….. anything else I can try?  thanks in advance!

                1 Reply Last reply Reply Quote 0
                • H
                  heper
                  last edited by

                  are you running multiple layer-3 over a single layer-2 ? if yes: don't

                  1 Reply Last reply Reply Quote 0
                  • J
                    JamesVA
                    last edited by

                    heper - it's [iperf client] <-> [pfsense VM] <-> [iperf server] that all sit on the same switch;

                    "iperf client", "pfsense VM" and "iperf server" are each on their own hardware.

                    I don't this would be considered as multiple L3 setup right?

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