Changes between Initial Version and Version 1 of TriagingTestFailures


Ignore:
Timestamp:
Nov 2, 2010 9:10:17 AM (13 years ago)
Author:
Adam Roben
Comment:

Initial version

Legend:

Unmodified
Added
Removed
Modified
  • TriagingTestFailures

    v1 v1  
     1This guide will help you triage test failures on build.webkit.org and get the bots back to being green.
     2
     3 1. Find out what is failing. There are two main ways to do this:
     4   * Browse build.webkit.org (you should probably start with this one)
     5     1. Find recent builds that have failed
     6       * Windows:
     7         1. Go to http://build.webkit.org/buildslaves.
     8         1. Click on the name of a slave you're interested in to see a summary of its recent builds.
     9       * Other platforms:
     10         1. Go to http://build.webkit.org/builders.
     11         1. Click on the name of a builder you're interested in to see a summary of its recent builds.
     12       * The '''Info''' column will tell you if any tests failed for that build.
     13       * To see the test output for a particular build, click on the link in the '''Build #''' column, then on '''view results''', then on '''results.html'''
     14   * Use `webkit-patch`
     15     1. Run this command:
     16{{{
     17webkit-patch failure-reason
     18}}}
     19     1. When prompted, specify which builder you're interested in.
     20     1. Press Enter to continue. `webkit-patch` will look back through the recent builds for that builder until it has found when all current failures were introduced.
     21 1. Group failures by "root causes"
     22   1. If a multiple tests are all failing for the same reason, they should be grouped together into a single root cause
     23 1. File one or more bugs for each root cause. (If a test fails on multiple platforms and those platforms will need separate fixes, you should file one bug for each failing platform.)
     24   1. Go to http://webkit.org/new-bug
     25   1. Include in your report:
     26     * The name(s) of the failing test(s)
     27     * What platform(s) the failures occur on
     28     * When the failures began, if known
     29     * A link to the failing output
     30       * If a single test had incorrect output, link to the pretty diff.
     31       * If multiple tests had incorrect output, or if the failure is a crash or hang, link to results.html.
     32   1. Apply keywords
     33     * `LayoutTestFailure`
     34     * `Regression`, if the failure is due to a regression in WebKit
     35     * `PlatformOnly`, if the test only fails on one platform
     36 1. Get the bots green again
     37   * If the tests always fail with the same incorrect output, check in new results for the tests and include the bug URL in your ChangeLog
     38   * If the tests fail intermittently, or crash, or hang, add the tests to the appropriate Skipped files. Include a comment in the Skipped file with the bug URL and a brief description of how it fails (e.g., "Crashes", "Sometimes times out", etc.).