Changes between Initial Version and Version 1 of PlasticSlices/Continuation/Round5


Ignore:
Timestamp:
03/07/12 10:51:07 (10 years ago)
Author:
Josh Smift
Comment:

--

Legend:

Unmodified
Added
Removed
Modified
  • PlasticSlices/Continuation/Round5

    v1 v1  
     1[[PageOutline]]
     2
     3= Continuation Round 5 =
     4
     5In this round, I ran plastic-105 and plastic-106 for about an hour, with only GT and Wisconsin, to try to track down some issues with performance in Round 3 that seemed related to GT. In particular, we wanted to find out whether there was any difference when GT was sending data vs receiving data.
     6
     7The raw logs are at http://www.gpolab.bbn.com/plastic-slices/continuation/round-5/.
     8
     9= plastic-105 =
     10
     11[ggw:PlasticSlices/Experiments#SteadyPerfTCP SteadyPerf TCP], using port=5105, time=518400, and this table of client/server pairs:
     12
     13|| '''client'''                   || '''server'''                   || '''server address''' ||
     14|| wings-openflow-2.wail.wisc.edu || plnode1.cip.gatech.edu         || server=10.42.105.100 ||
     15|| wings-openflow-3.wail.wisc.edu || plnode2.cip.gatech.edu         || server=10.42.105.101 ||
     16
     17== One-time prep commands run on each client and server ==
     18
     19{{{
     20sudo yum -y install iperf
     21}}}
     22
     23== Commands run on each server ==
     24
     25{{{
     26server=<ipaddr>
     27nice -n 19 iperf -B $server -p 5105 -s -i 1
     28}}}
     29
     30== Commands run on each client ==
     31
     32{{{
     33server=<ipaddr>
     34nice -n 19 iperf -c $server -p 5105 -t 518400
     35}}}
     36
     37== Results ==
     38
     39Generated with
     40
     41{{{
     42subnet=105
     43for host in $(awk 'NR%2==1' ~/plastic-slices/logins/logins-plastic-$subnet.txt | sed -r -e 's/.+@//') ; do echo -e "$host:\n\n{{{" ; grep -A 5 -B 1 "Client connecting" pgenigpolabbbncom_plastic$subnet\@$host.log ; echo -e "}}}\n" ; done
     44}}}
     45
     46and then edited slightly to remove artifacts (like control characters, my prompt, etc).
     47
     48wings-openflow-2.wail.wisc.edu:
     49
     50{{{
     51------------------------------------------------------------
     52Client connecting to 10.42.105.100, TCP port 5105
     53TCP window size: 16.0 KByte (default)
     54------------------------------------------------------------
     55[  3] local 10.42.105.95 port 55411 connected with 10.42.105.100 port 5105
     56[ ID] Interval       Transfer     Bandwidth
     57[  3]  0.0-4904.3 sec  1.19 GBytes  2.09 Mbits/sec
     58}}}
     59
     60wings-openflow-3.wail.wisc.edu:
     61
     62{{{
     63------------------------------------------------------------
     64Client connecting to 10.42.105.101, TCP port 5105
     65TCP window size: 16.0 KByte (default)
     66------------------------------------------------------------
     67[  3] local 10.42.105.96 port 52296 connected with 10.42.105.101 port 5105
     68[ ID] Interval       Transfer     Bandwidth
     69[  3]  0.0-4903.6 sec  1.18 GBytes  2.06 Mbits/sec
     70}}}
     71
     72== Analysis ==
     73
     74Not great numbers, but a factor of seven or eight times better than the other direction. This smells like a duplex problem to me -- remember that traffic is sent from the client to the server, in iperf.
     75
     76= plastic-106 =
     77
     78[ggw:PlasticSlices/Experiments#SteadyPerfUDP SteadyPerf UDP], using port=5106, time=518400, rate=30, and this table of client/server pairs:
     79
     80|| '''client'''                   || '''server'''                   || '''server address''' ||
     81|| wings-openflow-2.wail.wisc.edu || plnode2.cip.gatech.edu         || server=10.42.105.101 ||
     82|| wings-openflow-3.wail.wisc.edu || plnode1.cip.gatech.edu         || server=10.42.105.100 ||
     83
     84== One-time prep commands run on each client and server ==
     85
     86{{{
     87sudo yum -y install iperf
     88}}}
     89
     90== Commands run on each server ==
     91
     92{{{
     93server=<ipaddr>
     94nice -n 19 iperf -u -B $server -p 5106 -s -i 1
     95}}}
     96
     97== Commands run on each client ==
     98
     99{{{
     100server=<ipaddr>
     101nice -n 19 iperf -u -c $server -p 5106 -t 518400 -b 30M
     102}}}
     103
     104== Results ==
     105
     106Generated with
     107
     108{{{
     109subnet=106
     110for host in $(awk 'NR%2==1' ~/plastic-slices/logins/logins-plastic-$subnet.txt | sed -r -e 's/.+@//') ; do echo -e "$host:\n\n{{{" ; grep -A 10 -B 1 "Client connecting" pgenigpolabbbncom_plastic$subnet\@$host.log ; echo -e "}}}\n" ; done
     111}}}
     112
     113and then edited slightly to remove artifacts (like control characters, my prompt, etc).
     114
     115wings-openflow-2.wail.wisc.edu:
     116
     117{{{
     118------------------------------------------------------------
     119Client connecting to 10.42.105.101, UDP port 5106
     120Sending 1470 byte datagrams
     121UDP buffer size:  110 KByte (default)
     122------------------------------------------------------------
     123[  3] local 10.42.105.95 port 50439 connected with 10.42.105.101 port 5106
     124[ ID] Interval       Transfer     Bandwidth
     125[  3]  0.0-4900.4 sec  17.1 GBytes  30.0 Mbits/sec
     126[  3] Sent 12501122 datagrams
     127[  3] Server Report:
     128[  3]  0.0-4899.9 sec  17.1 GBytes  30.0 Mbits/sec   0.613 ms 6997/12501121 (0.056%)
     129[  3]  0.0-4899.9 sec  161 datagrams received out-of-order
     130}}}
     131
     132wings-openflow-3.wail.wisc.edu:
     133
     134{{{
     135------------------------------------------------------------
     136Client connecting to 10.42.105.100, UDP port 5106
     137Sending 1470 byte datagrams
     138UDP buffer size:  110 KByte (default)
     139------------------------------------------------------------
     140[  3] local 10.42.105.96 port 48014 connected with 10.42.105.100 port 5106
     141[ ID] Interval       Transfer     Bandwidth
     142[  3]  0.0-4903.0 sec  17.1 GBytes  30.0 Mbits/sec
     143[  3] Sent 12507660 datagrams
     144[  3] Server Report:
     145[  3]  0.0-4902.8 sec  17.1 GBytes  30.0 Mbits/sec   0.852 ms 9669/12507659 (0.077%)
     146[  3]  0.0-4902.8 sec  344 datagrams received out-of-order
     147}}}
     148
     149== Analysis ==
     150
     151Not perfect, but three nines isn't bad. A significant difference over the other direction, anyway.