lp://staging/~doanac/qa-dashboard/smoke-kpi-fix

Branch merges

Propose for merging

No branches dependent on this one.

Merged into lp://staging/qa-dashboard at revision 720

Chris Johnston: Approve on 2014-02-18

PS Jenkins bot: Approve (continuous-integration) on 2014-02-12

Diff: 66 lines (+8/-10)

3 files modified

smokeng/dashboard.py (+1/-1)
smokeng/management/commands/jenkins_pull_smokeng.py (+0/-9)
smokeng/models.py (+7/-0)

api

Related source package recipes

No recipes using this branch. (?)

Create packaging recipe

Related snap packages

Related bugs

Related blueprints

717. By Andy Doan on 2014-02-12

smoke kpi: fix calculation

now that we have results that can be "running" and "syncing" we need to
update what we query for. Additionally I noticed that "pass_rate" wasn't
being set for results that were "syncing". I could have fixed this from
api.py, but it seemed better to do this in a single place and then
simplify our pull-script logic.

716. By Chris Johnston on 2014-02-10

2013.02.10

715. By Andy Doan on 2014-02-10

[r=Chris Johnston] filter smoketestresult artifacts properly for mega jobs

The logic was already there, but was only used for SmokeResults 1278510 from Andy Doan

714. By Joe Talbot <email address hidden> on 2014-02-07

[r=Andy Doan] Smoke - Do not include test results for hidden images or results. 1277581 from Joe Talbott

713. By Andy Doan on 2014-02-06

[r=PS Jenkins bot, Joe Talbott] allow publishing of smoketestresults via REST API

Allow the client to include the UTAH YAML with its SmokeResult object
so that the SmokeTestResult details can also be populated
  from Andy Doan

712. By Chris Johnston on 2014-02-03

2014.02.03

711. By Joe Talbot <email address hidden> on 2014-02-03

[r=Andy Doan, PS Jenkins bot] Bootspeed - No longer require unique md5s.

* we aren't using md5s for anything and we've got a constraint on:
  - release, variant, arch, build_number from Joe Talbott

710. By Andy Doan on 2014-01-30

[r=Joe Talbott, PS Jenkins bot] fix artifact issues with mega-jobs

A jenkins job may now have artifacts that are specific to a testrun that occurred in a mega-job. As the code currently stands, a crash in one test will wind up being counted for all the test runs. This causes us to show 25 crashes instead of just 1. Additionally, we are showing log files for the job and not just for the run. This adds logic to show the right thing when we are dealing with a mega-job. from Andy Doan

709. By Chris Johnston on 2014-01-08

2014.01.08

708. By Joe Talbot <email address hidden> on 2014-01-08

[r=Andy Doan, PS Jenkins bot] smoke - Ignore publish=False entries when using get_or_create() from Joe Talbott