<div dir="auto">Thanks, let me evaluate this<div dir="auto"><br></div><div dir="auto">Bob</div></div><br><div class="gmail_quote"><div dir="ltr" class="gmail_attr">On Fri, Jan 24, 2020, 1:25 PM Toke Høiland-Jørgensen <<a href="mailto:toke@redhat.com">toke@redhat.com</a>> wrote:<br></div><blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex">Bob McMahon via Make-wifi-fast <<a href="mailto:make-wifi-fast@lists.bufferbloat.net" target="_blank" rel="noreferrer">make-wifi-fast@lists.bufferbloat.net</a>><br>
writes:<br>
<br>
> From: Bob McMahon <<a href="mailto:bob.mcmahon@broadcom.com" target="_blank" rel="noreferrer">bob.mcmahon@broadcom.com</a>><br>
> Subject: iperf 2.0.14<br>
> To: Make-Wifi-fast <<a href="mailto:make-wifi-fast@lists.bufferbloat.net" target="_blank" rel="noreferrer">make-wifi-fast@lists.bufferbloat.net</a>><br>
> Date: Fri, 24 Jan 2020 12:52:16 -0800<br>
><br>
> Hi All,<br>
><br>
> Just a heads up that iperf 2.0.14 <<a href="https://sourceforge.net/projects/iperf2/" rel="noreferrer noreferrer" target="_blank">https://sourceforge.net/projects/iperf2/</a>><br>
> is getting closer to a release. Feel free to suggest features that you<br>
> might find helpful.<br>
<br>
I have an old TODO item sitting around telling me to go over which<br>
features we use from netperf in Flent that we don't currently use for<br>
iperf. So thank you for the reminder; since I'm leaving for holidays<br>
tomorrow, here is a quick overview (i.e., I haven't had time to go over<br>
how many of these are actually missing from iperf, and how many are<br>
there, but we're just not using them):<br>
<br>
- We extract the following variables from netperf (-k option):<br>
<br>
output_vars = 'THROUGHPUT,LOCAL_CONG_CONTROL,REMOTE_CONG_CONTROL,' \<br>
'TRANSPORT_MSS,LOCAL_TRANSPORT_RETRANS,' \<br>
'REMOTE_TRANSPORT_RETRANS,LOCAL_SOCKET_TOS,' \<br>
'REMOTE_SOCKET_TOS,DIRECTION,ELAPSED_TIME,PROTOCOL,' \<br>
'LOCAL_SEND_SIZE,LOCAL_RECV_SIZE,' \<br>
'REMOTE_SEND_SIZE,REMOTE_RECV_SIZE,' \<br>
'LOCAL_BYTES_SENT,LOCAL_BYTES_RECVD,' \<br>
'REMOTE_BYTES_SENT,REMOTE_BYTES_RECVD'<br>
<br>
- We pass in DUMP_TCP_INFO to netperf and capture the output (ref:<br>
<a href="https://github.com/tohojo/flent/issues/13" rel="noreferrer noreferrer" target="_blank">https://github.com/tohojo/flent/issues/13</a>)<br>
<br>
- We use the netperf features to set:<br>
- diffserv markings<br>
- TCP congestion control<br>
- socket timeout<br>
<br>
- For running the public test servers, the ability to just leave a<br>
'netserver' instance running indefinitely and let it take care of<br>
negotiating test setups. I don't think this works that well with<br>
iperf?<br>
<br>
<br>
<br>
If iperf can replicate all of that, it'll be fairly straight-forward to<br>
make Flent backend-agnostic on the TCP test runner front. How much do<br>
you reckon is currently missing, and what would it take to get that in? :)<br>
<br>
-Toke<br>
<br>
_______________________________________________<br>
Make-wifi-fast mailing list<br>
<a href="mailto:Make-wifi-fast@lists.bufferbloat.net" target="_blank" rel="noreferrer">Make-wifi-fast@lists.bufferbloat.net</a><br>
<a href="https://lists.bufferbloat.net/listinfo/make-wifi-fast" rel="noreferrer noreferrer" target="_blank">https://lists.bufferbloat.net/listinfo/make-wifi-fast</a></blockquote></div>