From 6640a2ca7cc319e4b9957f074fd2f93eba649856 Mon Sep 17 00:00:00 2001 From: wym_libra Date: Thu, 21 Jan 2016 19:45:55 +0800 Subject: Test Result of TC019 and TC025 JIRA: None Change-Id: I425f957dade4bdb3bffe2f6c172fa6c3b68ce382 Signed-off-by: wym_libra --- docs/results/yardstick-opnfv-ha.rst | 94 ++++++++++++++++++++++++++++++++++--- 1 file changed, 87 insertions(+), 7 deletions(-) diff --git a/docs/results/yardstick-opnfv-ha.rst b/docs/results/yardstick-opnfv-ha.rst index 50eb414bb..4ee9de847 100644 --- a/docs/results/yardstick-opnfv-ha.rst +++ b/docs/results/yardstick-opnfv-ha.rst @@ -10,29 +10,109 @@ Test Results for yardstick-opnfv-ha .. toctree:: :maxdepth: 2 - Details ======= +There are two test cases, TC019 and TC025, for high availability (HA) test of +OPNFV platform, and both test cases were executed in CMCC's lab with 3+2 HA +deployment, where the installer is Arno SR1 release of fuel. -.. after this doc is filled, remove all comments and include the scenario in -.. results.rst by removing the comment on the file name. +TC019 +----- +This test case verifies the high availability of the openstack service, i.e. +"nova-api", on controller node. +There are one attacker, "kill-process" which kills all "nova-api" processes, +and two monitors, "openstack-cmd" monitoring "nova-api" service by openstack +command "nova image-list", while "process" monitor checks whether "nova-api" +process is running. Please see the test case description document for detail. Overview of test results ------------------------ +The service_outage_time of "nova image-list" is 0 seconds, while the +process_recover_time of "nova-api" is 300 seconds which equals the running time +of this test case, that means the "nova-api" service can't automatiocally +recover itself. + +Detailed test results +--------------------- +All "nova-api" process on the selected controller node was killed, and results +of two monitors were collected. Specifically, the results of "nova image-list" +request were collected from compute node and the status of "nova-api" process +were collected from the selected controller node. + +Each monitor was running in a single process. The running time of each monitor +was about 300 seconds with no waiting time between twice monitor running. For +"nova image-list", the running times is 127, that's to say there is one +openstack command request every 2.36 seconds; while the running times is 141 +for "nova-api" process checking, the accurancy is about 2.13 seconds. + +The outage time of each monitor, which the name is "service_outage_time" for +"openstack-cmd" monitor and "process_recover_time" for "process" monitor, is +defined as the duration from the begin time of the first failure request to the +end time of the last failure request. + +All "nova image-list" requestes were success, so the service_outage_time of +"nova image-list" is 0 second, while "nova-api" processes were not running for +all "process" checking, so the process_recover_time of "nova-api" is 300s. + +Rationale for decisions +----------------------- +The service_outage_time is 0 second, that means the failover time of openstack +service is less than 2.36s, which is the period of each request. However, the +process_recover_time equals test case runing time, that means the process is +not automatically recovered, so this test case is fail. + + +TC025 +----- +This test case verifies the high availability of controller node. When one of +the controller node abnormally shutdown, the service provided should be OK. +There are one attacker, "kill-process" which kills all "nova-api" processes, +and two "openstack-cmd" monitors, one monitoring openstack command +"nova image-list" and the other monitoring "neutron router-list". +Please see the test case description document for detail. -.. general on metrics collected, number of iterations +Overview of test results +------------------------ +The both service_outage_time of "nova image-list" and "neutron router-list" +were 0 second. Detailed test results --------------------- +A selected controller node was shutdown, and results of two monitors were +collected from compute node. -.. info on lab, installer, scenario +The return results of "nova image-list" and "neutron router-list" requests from +compute node were collected, then the failure requestion time were statistic +service_outage_time of corresponding service. + +Each monitor was running in a single process. The running time of each monitor +was about 300 seconds with no waiting time between twice monitor running. For +"nova image-list", the running times is 49, that's to say there is one +openstack command request every 6.12 seconds; while the running times is 28 for +"neutron router-list", the accurancy is about 10.71 seconds. + +The "service_outage_time" for two monitors is defined as the duration from the +begin time of the first failure request to the end time of the last failure +request. + +All "nova image-list" and "neutron router-list" requestes were success, so the +service_outage_time of both two monitor were 0 second. Rationale for decisions ----------------------- -.. result analysis, pass/fail +As service_outage_time of all monitors are 0 second, that means there are none +failure request in this test case running time, this test case is passed. + Conclusions and recommendations ------------------------------- +The TC019 shows the killed process will be not automatically recovered, which +should be imporved. -.. did the expected behavior occured? +There are several improvement points for HA test: +a) Running test cases in different enveriment deployed by different installers, +such as compass4nfv, apex and joid, with different versiones. +b) The period of each request is a little long, it needs more accurate test + method. +c) More test cases with different faults and different monitors are needed. -- cgit 1.2.3-korg