| 1 | [[PageOutline]] |
| 2 | |
| 3 | = Continuation Round 4 = |
| 4 | |
| 5 | In this round, we ran plastic-105 and plastic-106 for about three hours, 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 a specific problem with plnode1. |
| 6 | |
| 7 | The raw logs are at http://www.gpolab.bbn.com/plastic-slices/continuation/round-4/. |
| 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 | || plnode1.cip.gatech.edu || wings-openflow-2.wail.wisc.edu || server=10.42.105.95 || |
| 15 | || plnode2.cip.gatech.edu || wings-openflow-3.wail.wisc.edu || server=10.42.105.96 || |
| 16 | |
| 17 | == One-time prep commands run on each client and server == |
| 18 | |
| 19 | {{{ |
| 20 | sudo yum -y install iperf |
| 21 | }}} |
| 22 | |
| 23 | == Commands run on each server == |
| 24 | |
| 25 | {{{ |
| 26 | server=<ipaddr> |
| 27 | nice -n 19 iperf -B $server -p 5105 -s -i 1 |
| 28 | }}} |
| 29 | |
| 30 | == Commands run on each client == |
| 31 | |
| 32 | {{{ |
| 33 | server=<ipaddr> |
| 34 | nice -n 19 iperf -c $server -p 5105 -t 518400 |
| 35 | }}} |
| 36 | |
| 37 | == Results == |
| 38 | |
| 39 | Generated with |
| 40 | |
| 41 | {{{ |
| 42 | subnet=105 |
| 43 | for 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 | |
| 46 | and then edited slightly to remove artifacts (like control characters, my prompt, etc). |
| 47 | |
| 48 | plnode1.cip.gatech.edu: |
| 49 | |
| 50 | {{{ |
| 51 | ------------------------------------------------------------ |
| 52 | Client connecting to 10.42.105.95, TCP port 5105 |
| 53 | TCP window size: 16.0 KByte (default) |
| 54 | ------------------------------------------------------------ |
| 55 | [ 3] local 10.42.105.100 port 59778 connected with 10.42.105.95 port 5105 |
| 56 | [ 3] 0.0-13761.8 sec 405 MBytes 247 Kbits/sec |
| 57 | }}} |
| 58 | |
| 59 | plnode2.cip.gatech.edu: |
| 60 | |
| 61 | {{{ |
| 62 | ------------------------------------------------------------ |
| 63 | Client connecting to 10.42.105.96, TCP port 5105 |
| 64 | TCP window size: 16.0 KByte (default) |
| 65 | ------------------------------------------------------------ |
| 66 | [ 3] local 10.42.105.101 port 58242 connected with 10.42.105.96 port 5105 |
| 67 | [ 3] 0.0-13768.4 sec 482 MBytes 294 Kbits/sec |
| 68 | }}} |
| 69 | |
| 70 | == Analysis == |
| 71 | |
| 72 | Less than 300 Kbits/sec is very low. |
| 73 | |
| 74 | Looking at the server logs shows times when the rate jumped up to one or two Mbits/sec, before dropping back down to double-digt Kbits/sec. |
| 75 | |
| 76 | The results are similar for the two plnodes, so there probably isn't a problem with just one or the other. |
| 77 | |
| 78 | = plastic-106 = |
| 79 | |
| 80 | [ggw:PlasticSlices/Experiments#SteadyPerfUDP SteadyPerf UDP], using port=5106, time=518400, rate=30, and this table of client/server pairs: |
| 81 | |
| 82 | || '''client''' || '''server''' || '''server address''' || |
| 83 | || plnode1.cip.gatech.edu || wings-openflow-3.wail.wisc.edu || server=10.42.105.96 || |
| 84 | || plnode2.cip.gatech.edu || wings-openflow-2.wail.wisc.edu || server=10.42.105.95 || |
| 85 | |
| 86 | == One-time prep commands run on each client and server == |
| 87 | |
| 88 | {{{ |
| 89 | sudo yum -y install iperf |
| 90 | }}} |
| 91 | |
| 92 | == Commands run on each server == |
| 93 | |
| 94 | {{{ |
| 95 | server=<ipaddr> |
| 96 | nice -n 19 iperf -u -B $server -p 5106 -s -i 1 |
| 97 | }}} |
| 98 | |
| 99 | == Commands run on each client == |
| 100 | |
| 101 | {{{ |
| 102 | server=<ipaddr> |
| 103 | nice -n 19 iperf -u -c $server -p 5106 -t 518400 -b 30M |
| 104 | }}} |
| 105 | |
| 106 | == Results == |
| 107 | |
| 108 | Generated with |
| 109 | |
| 110 | {{{ |
| 111 | subnet=106 |
| 112 | for 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 |
| 113 | }}} |
| 114 | |
| 115 | and then edited slightly to remove artifacts (like control characters, my prompt, etc). |
| 116 | |
| 117 | plnode1.cip.gatech.edu: |
| 118 | |
| 119 | {{{ |
| 120 | ------------------------------------------------------------ |
| 121 | Client connecting to 10.42.105.96, UDP port 5106 |
| 122 | Sending 1470 byte datagrams |
| 123 | UDP buffer size: 122 KByte (default) |
| 124 | ------------------------------------------------------------ |
| 125 | [ 3] local 10.42.105.100 port 53151 connected with 10.42.105.96 port 5106 |
| 126 | [ 3] 0.0-13755.2 sec 48.0 GBytes 30.0 Mbits/sec |
| 127 | [ 3] Sent 35057619 datagrams |
| 128 | [ 3] Server Report: |
| 129 | [ 3] 0.0-13754.8 sec 44.2 GBytes 27.6 Mbits/sec 0.078 ms 2767576/35057619 (7.9%) |
| 130 | [ 3] 0.0-13754.8 sec 252 datagrams received out-of-order |
| 131 | }}} |
| 132 | |
| 133 | plnode2.cip.gatech.edu: |
| 134 | |
| 135 | {{{ |
| 136 | ------------------------------------------------------------ |
| 137 | Client connecting to 10.42.105.95, UDP port 5106 |
| 138 | Sending 1470 byte datagrams |
| 139 | UDP buffer size: 122 KByte (default) |
| 140 | ------------------------------------------------------------ |
| 141 | [ 3] local 10.42.105.101 port 39327 connected with 10.42.105.95 port 5106 |
| 142 | [ 3] 0.0-13756.4 sec 48.0 GBytes 30.0 Mbits/sec |
| 143 | [ 3] Sent 35068319 datagrams |
| 144 | [ 3] Server Report: |
| 145 | [ 3] 0.0-13756.0 sec 44.4 GBytes 27.7 Mbits/sec 0.252 ms 2666045/35068319 (7.6%) |
| 146 | [ 3] 0.0-13756.0 sec 319 datagrams received out-of-order |
| 147 | }}} |
| 148 | |
| 149 | == Analysis == |
| 150 | |
| 151 | Similar packet loss to Round 3, where we saw 7.5% loss. |
| 152 | |
| 153 | The results are similar for the two plnodes, so there probably isn't a problem with just one or the other. |