Hello All, I'm new to this mailing list, rather new to using Iperf, and I must warn you that my background is neither in programming or networking. I currently have Iperf installed on Windows XP machines and I'm curious about the behavior of the tool, but more specifically, the -w switch. Because of some of the issues I experienced using this Windows port of Iperf, I plan on using the native version on a Unix based OS, but I think my questions will apply then as well. I'm testing over a high-latency satellite link. I know this is long, please bare with me. Any help would be greatly appreciated.
I've read through the archives and online but still can't get a direct answer to m questions. Here's the most helpful info I've found in the archives: http://sourceforge.net/mailarchive/message.php?msg_name=469577EC.2040304 %40room52.net http://sourceforge.net/mailarchive/message.php?msg_name=4695C66E.5020205 %40room52.net This is my understanding so far, please correct me if (read: when) I'm wrong: (1) The -w switch really isn't the TCP Receive Window as the documentation says but rather the send/receive socket buffer size, and is not directly related to TCP Window Size. This goes against what I had originally assumed from the documentation and pretty much everything I've read online. (2) When the -w switch is used on the receiving server (ex: iperf -s -w <XX>), this should set the receive buffer size to <XX> (SO_RCVBUF). This is limited by some maximum value in the operating system. (3) When the -w switch is used on the sending client (ex: iperf -c -w <XX>), this should set the send buffer size to <XX> (SO_SNDBUF). This is limited by some maximum value in the operating system. Here are my questions/confusion: (1) Should the values set using the -w switch be shown in a packet capture such as Wireshark? When I use the -w switch on the server, the TCP Receive Window seems to remain unchanged and is usually 65535 bytes. I see the TCP Window change when I use -w values ranging from 32KB to 64KB. Values less than 32KB are not reflected in the packet capture, but still seem to have an effect on throughput. Not using the -w switch shows that the buffer size is 8KB, but it too is not reflected in a packet capture. Using this default value results in low throughput. As you can see, I am VERY confused! (2) How does the -w switch behave for multidirectional tests (-r and -d)? If "iperf -c <IP> -w <XX> -r" is used, does the <XX> become the Receive Buffer Size (SO_RCVBUF) when it switches to receive mode? Does the server's value then become the Send Buffer Size (SO_SNDBUF)? (3) So how does the socket buffer size actually relate to the TCP Receive Window? (4) Should the -w switch be used for UDP tests? (5) How should I use the -b (--bandwidth) switch for UDP tests? If I have a link bandwidth that is 128kbps and I know that I'm going to get a lower "goodput" due to overhead (IP, TCP, IPSec, tunneling), do I set the bandwidth to -b 128000, a lesser value, or am I completely misunderstanding the concept? Many Many Thanks, Jimmy El Zorkani ------------------------------------------------------------------------------ This SF.net Dev2Dev email is sponsored by: Show off your parallel programming skills. Enter the Intel(R) Threading Challenge 2010. http://p.sf.net/sfu/intel-thread-sfd _______________________________________________ Iperf-users mailing list Iperf-users@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/iperf-users
