1 .. This work is licensed under a Creative Commons Attribution 4.0 International
3 .. http://creativecommons.org/licenses/by/4.0
6 ==============================================
7 Test Results for compass-os-nosdn-nofeature-ha
8 ==============================================
17 .. _Grafana: http://130.211.154.108/grafana/dashboard/db/yardstick-main
18 .. _SC_POD: https://wiki.opnfv.org/pharos?&#community_test_labs
20 Overview of test results
21 ------------------------
23 See Grafana_ for viewing test result metrics for each respective test case. It
24 is possible to chose which specific scenarios to look at, and then to zoom in
25 on the details of each run test scenario as well.
27 All of the test case results below are based on 5 consecutive scenario test
28 runs, each run on the Huawei SC_POD_ between February 13 and 18 in 2016. The
29 best would be to have more runs to draw better conclusions from, but these are
30 the only runs available at the time of OPNFV R2 release
34 The round-trip-time (RTT) between 2 VMs on different blades is measured using
35 ping. The measurements are on average varying between 1.95 and 2.23 ms
36 with a first 2 - 3.27 ms RTT spike in the beginning of each run (This could be
37 because of normal ARP handling).SLA set to 10 ms. The SLA value is used as a
38 reference, it has not been defined by OPNFV.
42 The IO read bandwidth look similar between different test runs, with an
43 average at approx. 145-162 MB/s. Within each run the results vary much,
44 minimum 2MB/s and maximum 712MB/s on the totality.
45 SLA set to 400KB/s. The SLA value is used as a reference, it has not been
50 The measurements for memory latency are consistent among test runs and results
51 in approx. 1.2 ns. The variations between runs are similar, between
52 1.215 and 1.278 ns. SLA set to 30 ns. The SLA value is used as
53 a reference, it has not been defined by OPNFV.
57 For this scenario no results are available to report on. Probable reason is
58 an integer/floating point issue regarding how InfluxDB is populated with
59 result data from the test runs.
63 The average measurements for memory bandwidth are consistent among most of the
64 different test runs at 12.98 - 16.73 GB/s. The last test run averages at
65 16.67 GB/s. Within each run the results vary, with minimal BW of 16.59
66 GB/s and maximum of 16.71 GB/s of the totality.
67 SLA set to 15 GB/s. The SLA value is used as a reference, it has not been
72 The Unixbench processor single and parallel speed scores show similar results
73 at approx. 3000. The runs vary between scores 2499 and 3105.
78 The round-trip-time (RTT) between VM1 with ipv6 router on different blades is
79 measured using ping6. The measurements are consistent at approx. 4 ms.
80 SLA set to 30 ms.The SLA value is used as a reference, it has not been
85 The amount of packets per second (PPS) and round trip times (RTT) between 2 VMs
86 on different blades are measured when increasing the amount of UDP flows sent
87 between the VMs using pktgen as packet generator tool.
89 Round trip times and packet throughput between VMs are typically affected by
90 the amount of flows set up and result in higher RTT and less PPS
93 When running with less than 10000 flows the results are flat and consistent.
94 RTT is then approx. 30 ms and the number of PPS remains flat at approx.
95 230000 PPS. Beyond approx. 10000 flows and up to 1000000 (one million) there
96 is an even drop in RTT and PPS performance, eventually ending up at approx.
97 105-113 ms and 100000 PPS respectively.
101 test purpose is to verify the function of Yang-to-Tosca in Parse, and this test
102 case is a weekly task, so it was triggered by manually, the result whether the
103 output is same with expected outcome is success
106 Detailed test results
107 ---------------------
109 The scenario was run on Huawei SC_POD_ with:
114 No SDN controller installed
116 Rationale for decisions
117 -----------------------
120 Tests were successfully executed and metrics collects (apart from TC011_).
121 No SLA was verified. To be decided on in next release of OPNFV.
123 Conclusions and recommendations
124 -------------------------------
126 The pktgen test configuration has a relatively large base effect on RTT in
127 TC037 compared to TC002, where there is no background load at all (30 ms
128 compared to 1 ms or less, which is more than a 3000 percentage different
129 in RTT results). The larger amounts of flows in TC037 generate worse
130 RTT results, in the magnitude of several hundreds of milliseconds. It would
131 be interesting to also make and compare all these measurements to completely
132 (optimized) bare metal machines running native Linux with all other relevant
133 tools available, e.g. lmbench, pktgen etc.