[netperf-talk] Inter-burst time
TC Carrington
tc_carrington at yahoo.com
Thu May 21 13:13:53 PDT 2009
I recently enabled intervals with the ./configure --enable-intervals call and added -DINTERVALS to the CFlags section of the makefile and recompiled after running a make clean. The problem I'm having is I when I use the -w global option on the TCP_STREAM test I get the same throughputs, at least within error, regardlesss of how large or small I set the inter-burst time to. It lets me run the test, adds ":interval" to the end, and runs through without error. When I enable debugging everything looks good except the signal catcher:
[root at ralph ~]# netperf -H 192.168.100.13 -d -w 100
scan_sockets_args called with the following argument vector
netperf -H 192.168.100.13 -d -w 100
Program name: netperf
Local send alignment: 8
Local recv alignment: 8
Remote send alignment: 8
Remote recv alignment: 8
Report local CPU 0
Report remote CPU 0
Verbosity: 1
Debug: 1
Port: 12865
Test name: TCP_STREAM
Test bytes: 0 Test time: 10 Test trans: 0
Host name: 192.168.100.13
installing catcher for all signals
Could not install signal catcher for sig 9, errno 22
Could not install signal catcher for sig 19, errno 22
Could not install signal catcher for sig 32, errno 22
Could not install signal catcher for sig 33, errno 22
Could not install signal catcher for sig 65, errno 22
remotehost is 192.168.100.13 and port 12865
establish_control called with host '192.168.100.13' port '12865' remfam AF_UNSPE C
local '0.0.0.0' port '0' locfam AF_UNSPEC
getaddrinfo returned the following for host '192.168.100.13' port '12865' family AF_UNSPEC
cannonical name: '192.168.100.13'
flags: 2 family: AF_INET: socktype: SOCK_STREAM protocol IPPROTO_TCP add rlen 16
sa_family: AF_INET sadata: 50 65 192 168 100 13
getaddrinfo returned the following for host '0.0.0.0' port '0' family AF_UNSPEC
cannonical name: '0.0.0.0'
flags: 3 family: AF_INET: socktype: SOCK_STREAM protocol IPPROTO_TCP add rlen 16
sa_family: AF_INET sadata: 0 0 0 0 0 0
bound control socket to 0.0.0.0 and 0
successful connection to remote netserver at 192.168.100.13 and 12865
complete_addrinfo using hostname 192.168.100.13 port 0 family AF_UNSPEC type SOC K_STREAM prot IPPROTO_TCP flags 0x0
getaddrinfo returned the following for host '192.168.100.13' port '0' family AF_ UNSPEC
cannonical name: '192.168.100.13'
flags: 2 family: AF_INET: socktype: SOCK_STREAM protocol IPPROTO_TCP add rlen 16
sa_family: AF_INET sadata: 0 0 192 168 100 13
local_data_address not set, using local_host_name of '0.0.0.0'
complete_addrinfo using hostname 0.0.0.0 port 0 family AF_UNSPEC type SOCK_STREA M prot IPPROTO_TCP flags 0x1
getaddrinfo returned the following for host '0.0.0.0' port '0' family AF_UNSPEC
cannonical name: '0.0.0.0'
flags: 3 family: AF_INET: socktype: SOCK_STREAM protocol IPPROTO_TCP add rlen 16
sa_family: AF_INET sadata: 0 0 0 0 0 0
TCP STREAM TEST from 0.0.0.0 (0.0.0.0) port 0 AF_INET to 192.168.100.13 (192.168 .100.13) port 0 AF_INET : interval
create_data_socket: socket 4 obtained...
netperf: set_sock_buffer: send socket size determined to be 16384
netperf: set_sock_buffer: receive socket size determined to be 87380
send_tcp_stream: send_socket obtained...
recv_response: received a 256 byte response
remote listen done.
About to start a timer for 10 seconds.
recv_response: received a 256 byte response
remote results obtained
calculate_confidence: itr 1; time 10.006981; res 1021.586483
lcpu -1.000000; rcpu -1.000000
lsdm -1.000000; rsdm -1.000000
Recv Send Send
Socket Socket Message Elapsed
Size Size Size Time Throughput
bytes bytes bytes secs. 10^6bits/sec
87380 16384 16384 10.01 1021.59
shutdown_control: shutdown of control connection requested.
When I increase the inter-burst time shouldn't the throughput go down? Also When I run a tcpdump the number of packets sent remain basically the same. Where am I going wrong here?
TC
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://www.netperf.org/pipermail/netperf-talk/attachments/20090521/e0d10399/attachment.htm
More information about the netperf-talk
mailing list