Quality Assurance SIG/20170421

2017-04-21

Attendees: Anthony, Elena, Erik, Greg, JR

Anthony:

it can be difficult the when and where of things will be testable. Another challenge is that the various testing environments don't have the necessary data/configuration to test.

Elena:

Testing Ores is going well, are planning to push/switch to en wiki. Testing in Polish wiki was a real goldmine.

Eirk:

Lots of learning as they test mjolnir. Testing is a little slow due to the nature of the test environment vs production, but performance isn't the focus in this test. Moving over to the analytics cluster when more production like testing is needed/desired.

JR:

Discussed progress/work in the release pipeline work that's going on. Containerized pipeline shows promise of providing parity between testing environments. Right now the primary work is around delivering a POC and requirements gathering. Elena, brought up the fact that parity with production is more complicated as the various wikis have significant differences. In addition to those differences, we will also need to address the test data gap.

JR asked that tribe provide requirements for the pipeline if they have any come to mind.

JR also mentioned the discussions that he's been having with people across the organization and the development of a summary of learnings and prospective changes that we could make. Will share out with the Tribe after the RelEng review meeting next week.

Wrap-up:

QA Tribe format, cadence, and participants to remain the same for now.