<div>Thanks, I will download from the trunk and run the test.</div> <div><BR><BR><B><I>Rick Jones <rick.jones2@hp.com></I></B> wrote:</div> <BLOCKQUOTE class=replbq style="PADDING-LEFT: 5px; MARGIN-LEFT: 5px; BORDER-LEFT: #1010ff 2px solid">anil mishra wrote:<BR>> Yes, having the header information is helpful to understand the output.<BR><BR>I wasn't going to remove the header, it just might not look the same.<BR><BR>> For TCP_RR, if it gives the calculated latency (RTT/2), that will be <BR>> helpful.<BR>> <BR>> Also, can it be possible to specify the mtu size, TCP Window parameter <BR>> etc. in the command if user wants to use it like that?<BR><BR>Netperf will never specify the MTU size. That is too far down the stack <BR>for netperf to be messing with. At present, it can _report_ the *MSS* <BR>of the connection when the verbosity is >= 2. I'm still waffling on <BR>whether or not to include the option to try to _set_ the MSS via the
<BR>TCP_MAXSEG option.<BR><BR>The TCP window settings are presumed to derive from the socket buffer <BR>settings made (-s and -S) before the connection is established.<BR><BR><BR>I have checked-in a first pass at some changes to have the TCP_RR test <BR>allow one to ask for bitrate rather than transactions per second as the <BR>primary figure of merit. Also, for the verbosity > 2 level, it will <BR>report the average RTT latensy, the transactions/s and the inbound and <BR>outbound bitrates.<BR><BR>You can try-out the bits from the top of trunk in the repository:<BR><BR>http://www.netperf.org/svn/netperf2/trunk/<BR><BR>From the Release Notes (not in the docs since this is still experimental)<BR><BR><EXCERPT><BR><BR>*) One can now pass a value of 'x' to the global -f option to specify<BR>the units as transactions per second. This is the default for any<BR>request/response test, which is determined by there being a "double<BR>`r'" in the name - eg "RR," "rr," "Rr," or
"rR." At present only<BR>the TCP_RR test actually looks for this to be set.<BR><BR>*) One can request bits/bytes per second as the primary output of a<BR>TCP_RR test by setting the global -f option to [kmgKMG] as with any<BR>of the "STREAM" tests. This converts the primary throughput metric<BR>to a bitrate (byterate) following the verbosity rules for a STREAM<BR>test. Service demand remains usec/Transaction regardless of the<BR>setting of the global -f option.<BR><BR>A verbosity level of 2 or more will cause the TCP_RR test to report<BR>calculated average RTT latency, transaction rate, and inbound and<BR>outbound transfer rates regardless of the primary units selected<BR>with the global -f paramter. If the primary output is transactions<BR>per second, the reported inbound and outbound transfer rates will<BR>be 10^6 bits per second, otherwise, they honor the setting of the<BR>global -f option.<BR><BR>All of this is EXPERIMENTAL and subject to change without prior<BR>notice
in future versions of netperf.<BR><BR></EXCERPT><BR><BR>an example with the primary figure of merit still transactions/s:<BR><BR>raj@tardy:~/netperf2_trunk$ src/netperf -H tardy.cup.hp.com -v 2 -f x -t <BR>TCP_RR -- -r 1024 -b 120 -S 1M -s 1M -D<BR>TCP REQUEST/RESPONSE TEST from 0.0.0.0 (0.0.0.0) port 0 AF_INET to <BR>tardy.cup.hp.com (15.244.56.217) port 0 AF_INET : nodelay : first burst 120<BR>Local /Remote<BR>Socket Size Request Resp. Elapsed Trans.<BR>Send Recv Size Size Time Rate<BR>bytes Bytes bytes bytes secs. per sec<BR><BR>219136 219136 1024 1024 10.00 9209.07<BR>1048576 1048576<BR>Alignment Offset RoundTrip Trans Throughput<BR>Local Remote Local Remote Latency Rate 10^6bits/s<BR>Send Recv Send Recv usec/Tran per sec Outbound Inbound<BR>8 0 0 0 13139.216 9209.073 75.441 75.441<BR><BR>and one with it as MB/s:<BR><BR>raj@tardy:~/netperf2_trunk$ src/netperf -H tardy.cup.hp.com -v 2 -f M -t <BR>TCP_RR -- -r 1024 -b 120 -S 1M -s 1M -D<BR>TCP REQUEST/RESPONSE TEST from
0.0.0.0 (0.0.0.0) port 0 AF_INET to <BR>tardy.cup.hp.com (15.244.56.217) port 0 AF_INET : nodelay : first burst 120<BR>Local /Remote<BR>Socket Size Request Resp. Elapsed<BR>Send Recv Size Size Time Throughput<BR>bytes Bytes bytes bytes secs. MBytes/sec<BR><BR>219136 219136 1024 1024 10.01 15.84<BR>1048576 1048576<BR>Alignment Offset RoundTrip Trans Throughput<BR>Local Remote Local Remote Latency Rate MBytes /s<BR>Send Recv Send Recv usec/Tran per sec Outbound Inbound<BR>8 0 0 0 14921.124 8109.309 7.919 7.919<BR><BR>note that both were using the "first burst" option. Here is a simpler <BR>one with just the classic single-byte request/response and no actual burst:<BR><BR>raj@tardy:~/netperf2_trunk$ src/netperf -H tardy.cup.hp.com -v 2 -t TCP_RR<BR>TCP REQUEST/RESPONSE TEST from 0.0.0.0 (0.0.0.0) port 0 AF_INET to <BR>tardy.cup.hp.com (15.244.56.217) port 0 AF_INET : first burst 0<BR>Local /Remote<BR>Socket Size Request Resp. Elapsed Trans.<BR>Send Recv Size Size Time
Rate<BR>bytes Bytes bytes bytes secs. per sec<BR><BR>16384 87380 1 1 10.00 783.88<BR>32768 32768<BR>Alignment Offset RoundTrip Trans Throughput<BR>Local Remote Local Remote Latency Rate 10^6bits/s<BR>Send Recv Send Recv usec/Tran per sec Outbound Inbound<BR>8 0 0 0 1275.702 783.882 0.006 0.006<BR><BR><BR>> <BR>> */Rick Jones <RICK.JONES2@HP.COM>/* wrote:<BR>> <BR>> If you depend heavily on the netperf header format, please let me know<BR>> ASAP. Since it is possible now to use a netperf TCP_RR test to perform<BR>> a bidirectional transfer on the same connection, I would like to modify<BR>> things so one can ask for output in the same units as the STREAM tests.<BR>> <BR>> This _may_ alter what one sees by default for the headers of a TCP_RR<BR>> test and I want to know if it will break anyone's scripts or whatnot.<BR>> <BR>> happy benchmarking,<BR>> <BR>> rick jones<BR>> _______________________________________________<BR>>
netperf-dev mailing list<BR>> netperf-dev@netperf.org<BR>> http://www.netperf.org/cgi-bin/mailman/listinfo/netperf-dev<BR>> <BR>> <BR>> ------------------------------------------------------------------------<BR>> Get the Yahoo! toolbar and be alerted to new email <BR>> <HTTP: evt="48225/*http://new.toolbar.yahoo.com/toolbar/features/mail/index.php" us.rd.yahoo.com>wherever <BR>> you're surfing.<BR><BR></BLOCKQUOTE><BR><p> 
<hr size=1>Need a vacation? <a href="http://us.rd.yahoo.com/evt=48256/*http://travel.yahoo.com/;_ylc=X3oDMTFhN2hucjlpBF9TAzk3NDA3NTg5BHBvcwM1BHNlYwNncm91cHMEc2xrA2VtYWlsLW5jbQ--">Get great deals
to amazing places </a>on Yahoo! Travel.