#opnfv-testperf: OPNFV functest

Meeting summary

1. 1. Jose Lausuch (jose_lausuch, 16:01:28)
   2. David Karr (davidmichaelkarr, 16:01:30)
   3. Malla Reddy (Malla, 16:01:49)
   4. Morgan Richomme (morgan_orange, 16:02:34)
   5. https://global.gotomeeting.com/join/566165845 (jose_lausuch, 16:04:02)
   
   
2. Status on Functest R1 (jose_lausuch, 16:07:43)
   1. Morgan going through the functest history and the existing test cases and tools (jose_lausuch, 16:13:40)
   2. https://global.gotomeeting.com/join/566165845 (jose_lausuch, 16:18:24)
   3. test projects have to think about how to present their results according to a common API (jose_lausuch, 16:34:08)
   
   
3. Results collection (jose_lausuch, 16:34:48)
   1. artifacts is not the right place to push the test results (jose_lausuch, 16:35:07)
   2. MongoDB will be used to store the results (jose_lausuch, 16:38:03)
   3. Releng will be the place to define the DB structure (jose_lausuch, 16:38:18)
   4. We don't know yet where to deploy the DB and where to store the results, we will wait for feedback from LF (jose_lausuch, 16:38:48)
   
   
4. Test coverage (jose_lausuch, 16:39:06)
   1. new wiki with R2 vision for testing strategy (jose_lausuch, 16:39:38)
   2. https://wiki.opnfv.org/testing (jose_lausuch, 16:39:41)
   3. table showing the testing category and to which testing project it belongs to (jose_lausuch, 16:42:23)
   4. new requirement projects to test OAM category, like Scalator (jose_lausuch, 16:43:12)
   5. table per domain (Compute, Storage, Network) according to the blocks in the ETSI point of view https://www.opnfv.org/sites/opnfv/files/blogs/opnfv_diagram_final.jpg (jose_lausuch, 16:49:00)
   6. some components will be out of scope (computing hw, storage hw, …) (jose_lausuch, 16:49:45)
   7. idea is to test that the system is cable of deploying vIMS as example of NFV, but not testing in depth vIMS (jose_lausuch, 16:52:25)
   8. scope is not to test all the NFV, not to have a catalog of all the NFV been tested. Reference will be vIMS (jose_lausuch, 16:53:05)
   9. David points out that having a candidate/reference VNF validated and tested is the point of opnfv, since it is a platform to deploy VNFs (jose_lausuch, 16:56:14)
   10. Malla says we should know where the application (VNF) comes from (jose_lausuch, 16:57:10)
   11. an opensource vIMS would be the reference to prove that the infrastructure is capable of deploying VNFs (jose_lausuch, 16:57:42)
   12. demo to be organized about functest R1 (jose_lausuch, 17:03:45)

Action items

1. (none)

People present (lines said)

1. jose_lausuch (37)
2. morgan_orange (6)
3. collabot (3)
4. Malla (2)
5. fdegir (2)
6. tfherbert1 (2)
7. davidmichaelkarr (1)