Dealing with Verification Data Overload
summary
Describes how AdoptOpenJDK and Eclipse OpenJ9 handle huge volumes of test logs and console output from many tests, versions, platforms and builds by applying data visualization, filtering and categorization to refine raw verification data. Mentions reducing verbosity via instrumentation of test code and experimenting with technologies such as deep learning to help display results efficiently, let users quickly narrow problems, and improve farm monitoring for complex continuous delivery pipelines.
AI-written summary · about
chapters (26)
- 0:00Intro
- 1:52The Scope
- 2:58Degrees of Freedom
- 4:44Gather Great Tests
- 5:16AdoptOpenJDK Quality Assurance (AQA)
- 6:10AQA Manifesto
- 8:26Test Framework (TestKitGen)
- 9:02Grouping & Granularity
- 9:34AdoptOpenJDK CI Pipeline
- 10:23Jenkins scripts for testing
- 11:02TAP & JUnit Plugin
- 11:30Archive data
- 11:53Test Result Summary Service (TRSS)
- 12:56TRSS Overview
- 13:30TRSS: Personalized Dashboard
- 13:54TRSS: Monitor Jenkins Pipeline Builds
- 14:22TRSS: Test Builds Result
- 14:40TRSS: Tests Result
- 15:36TRSS: Search test
- 15:49TRSS: Perf Dashboard
- 16:05Let Us Count the Ways
- 16:23What is Deep Learning?
- 17:09Initial DL Experiments
- 18:05Model Building
- 18:55Plans Forward
- 20:36Connect & Get Involved
related talks
21mDealing with Verification Data Overload
50mData-Driven Engineering
36mFrom Sledgehammer to Fine Brush for QA
29mJava Performance Testing for Everyone
50mPerformance Testing Java Applications
38mSetting up data driven tests with Java tools
35mSmarten Up Your Pipeline and Reduce Test Execution Time｜
49mHow to Optimize Automated Testing with Everyone's Favorite Butler