Version 74 (modified by 3 years ago) ( diff ) | ,
---|
The Jenkins continuous integration environment at https://djangoci.com builds pull requests using the GitHub Integration Plugin.
Commands
The main pull request builder is:
- pull-requests-bionic (Ubuntu 18.04) runs PostgreSQL 11, MySQL 5.7 and MariaDB 10.4.24. GIS: spatialite 4.3.0a-5, GDAL 2.2.3, GEOS 3.6.2, PostGIS 2.5.
Other pull request builders which can be triggered manually with the corresponding comment phrases:
- pull-requests-oracle runs on Oracle 19.3.0.0.0. The Oracle tests take about 40 minutes, so we avoid them unless we suspect a change may have broken something on Oracle.
buildbot, test on oracle.
- pull-requests-selenium runs the selenium tests.
buildbot, test on selenium.
- Retired:
- Ubuntu 12.04 runs PostgreSQL 9.1 and MySQL 5.5. GIS: spatialite 3.0.0~beta, GDAL 1.7.3, GEOS 3.2.2, PostGIS 1.5.3
- Ubuntu 14.04 runs PostgreSQL 9.3 and MySQL 5.6. GIS: spatialite 4.1.1, GDAL 1.10.1, GEOS 3.4.2, PostGIS 2.1
- Ubuntu 16.04 runs PostgreSQL 9.5 and MySQL 5.7. GIS: spatialite 4.3.0a-5, GDAL 1.11.3, GEOS 3.5.0, PostGIS 2.2
GitHub actions
Some other builders run on all pull requests via GitHub actions:
black
- runsblack
from the root checkout directory and fails if there are any warnings.flake8
- runsflake8
from the root checkout directory and fails if there are any warnings.docs
- runsmake spelling
from thedocs/
directory and fails if there are any spelling or sphinx errors.isort
- runsisort --check --diff .
from the root checkout directory and fails if any imports aren't sorted properly.JavaScript tests
- runs npm test and fails if there are any errorsWindows tests
- runs with Python 3.10.x.
Troubleshooting
Q: The build result disappeared. How do I get it back?
A: Build results are kept for a few days. Repush to your branch if you need to run the build again.
Q: The build failed with "ModuleNotFoundError: No module named 'memcache'"
A: The Python dependencies probably failed to install from PyPI. There's an issue where PyPI sometimes fails to return a correct response.
Q: Why is the disk nearly full on the Jenkins master node?
A: /var/log/jenkins/jenkins.log
may be many gigabytes in size. Remove that file and restart Jenkins with service jenkins restart
.