Classify tempest-devstack failures using ElasticSearch
Go to file
2013-11-26 19:07:44 +00:00
doc/source Apply Cookiecutter to the repo. 2013-09-23 15:27:39 -07:00
elastic_recheck Merge "Add launchpad support to check_success" 2013-11-26 19:07:44 +00:00
.coveragerc Apply Cookiecutter to the repo. 2013-09-23 15:27:39 -07:00
.gitignore Apply Cookiecutter to the repo. 2013-09-23 15:27:39 -07:00
.gitreview Apply Cookiecutter to the repo. 2013-09-23 15:27:39 -07:00
.testr.conf Apply Cookiecutter to the repo. 2013-09-23 15:27:39 -07:00
babel.cfg Apply Cookiecutter to the repo. 2013-09-23 15:27:39 -07:00
CONTRIBUTING.rst Apply Cookiecutter to the repo. 2013-09-23 15:27:39 -07:00
elasticRecheck.conf.sample The key's value is misleading in the config sample. 2013-10-18 19:57:21 -07:00
LICENSE Apply Cookiecutter to the repo. 2013-09-23 15:27:39 -07:00
MANIFEST.in Apply Cookiecutter to the repo. 2013-09-23 15:27:39 -07:00
queries.yaml Merge "Remove queries for dead bugs" 2013-11-26 19:07:43 +00:00
README.rst Add doc on queries.yaml 2013-11-19 20:40:37 -08:00
recheckwatchbot.yaml Make bot.py behave like a daemon 2013-09-18 17:45:12 -04:00
requirements.txt Make pid file configurable 2013-09-30 10:29:32 -07:00
setup.cfg Add graph script 2013-10-02 14:56:49 -07:00
setup.py Apply Cookiecutter to the repo. 2013-09-23 15:27:39 -07:00
test-requirements.txt Add mox fixture to base TestCase 2013-10-01 18:05:33 -04:00
tox.ini Reorganize tests into unit and functional tests 2013-10-09 13:52:25 -04:00

elastic-recheck

"Classify tempest-devstack failures using ElasticSearch"

Idea

When a tempest job failure is detected, by monitoring gerrit (using gerritlib), a collection of logstash queries will be run on the failed job to detect what the bug was.

Eventually this can be tied into the rechecker tool and launchpad

queries.yaml

All queries are stored in a yaml file called: queries.yaml

Guidelines for queries.yaml

  • After a bug is resolved and has no more hits in elasticsearch, it should be removed
  • Queries should get as close as possible to fingerprinting the root cause
  • Queries should not return any hits for successful jobs, this is a sign the query isn't specific enough

Future Work

  • Move config files into a separate directory
  • Make unit tests robust
  • Merge both binaries
  • Add debug mode flag
  • Split out queries repo
  • Expand gating testing
  • Cleanup and document code better
  • Move away from polling ElasticSearch to discover if its ready or not
  • Add nightly job to propose a patch to remove bug queries that return no hits -- Bug hasn't been seen in 2 weeks and must be closed

Main Dependencies

  • gerritlib
  • pyelasticsearch