Based on Arno, related to the result collection API could look like: * test-projects.json: description of the tests projects * test-functest.json: description of functest test project * test-functest-vping.json: raw result of 3 executions of vPing on 2 PODs (LF1 and LF2) The projects are described as follow in the database: http://213.77.62.197:80/test_projects { "meta": { "total": 1, "success": true }, "test_projects": [ { "test_cases": 0, "_id": "55e05b08514bc52914149f2d", "creation_date": "2015-08-28 12:58:48.602898", "name": "functest", "description": "OPNFV Functional testing project" } ]} If you want to know the testcases of this project http://213.77.62.197:80/test_projects/functest/cases { "meta": { "total": 1, "success": true }, "test_cases": [ { "_id": "55e05dba514bc52914149f2e", "creation_date": "2015-08-28 13:10:18.478986", "name": "vPing", "description": "This test consist in send exchange PING requests between two created VMs" } ]} Functest used to run automatically 4 suites: * vPing * ODL * Rally * Rally The results are available on [[https://build.opnfv.org/ci/view/functest/|Jenkins]] Functest calls the API to store the raw results of each tests. curl -X POST -H "Content-Type: application/json" -d '{"details": {"timestart": 123456, "duration": 66, "status": "OK"}, "project_name": "functest", "pod_id": 1, "case_name": "vPing"}' http://213.77.62.197:80/results From a Python client the API can be invoked as follow: def push_results_to_db(payload): url = TEST_DB params = {"project_name": "functest", "case_name": "vPing", "pod_id": 1, "details": payload} headers = {'Content-Type': 'application/json'} r = requests.post(url, data=json.dumps(params), headers=headers) logger.debug(r) with the payload var {'timestart': start_time_ts, 'duration': duration,'status': test_status}) Please note that that can put whatever you want in your payload. For vPing we put the useful information, for Rally we can directly put all the json report already built by Rally ([[http://artifacts.opnfv.org/functest/lf_pod1.localdomain/2015-06-18_06-09-20/results/rally/opnfv-authenticate.json|json result for Rally opnfv-authenticate scenario]]) In the DB you can see the results http://213.77.62.197:80/results?projects=functest&case=vPing { "meta": { "total": 17 }, "test_results": [ { "project_name": "functest", "description": null, "creation_date": "2015-08-28 16:43:00.965000", "case_name": "vPing", "details": { "timestart": 123456, "duration": 66, "status": "OK" }, "_id": "55e08f94514bc52b1791f949", "pod_id": 1 }, .......... { "project_name": "functest", "description": null, "creation_date": "2015-09-04 07:17:51.239000", "case_name": "vPing", "details": { "timestart": 1441350977.858555, "duration": 84.5, "status": "OK" }, "_id": "55e9459f514bc52b1791f959", "pod_id": 1 } ]} The raw results will be stored in the Mongo DB (json files) A module will perform post processing (cron) to generate json file usable for dashboarding (like in bitergia). In the exemple we build json file for LF2 POD (we could do it for the last 30 days, last 365 days, since the beginning) several files shall be produced according to the filters For vPing we shall produce: * result-functest-vPing----.json with * date = 30 (days), 365 (days), since the begining of the project * POD = LF-test1, Dell-test2, Orange-test1,... * installer = foreman, fuel, joid * version = stable Arno, Arno SR1, Stable Brahmaputra, ... The result-functest-vPing.json (all installers/POD/version since first day of the project) could look like: { { "description": "vPing results for Dashboard" }, { "info": { "xlabel": "time", "type": "graph", "ylabel": "duration" }, "data_set": [ { "y": 66, "x": "2015-08-28 16:43:00.965000" }, ................. { "y": 84.5, "x": "2015-09-04 07:17:51.239000" } ], "name": "vPing duration" }, { "info": { "type": "bar" }, "data_set": [ { "Nb Success": 17, "Nb tests": 17 } ], "name": "vPing status" }}