This functionality is going to be covered by packetary
https://github.com/openstack/fuel-mirror/tree/master/packetary
which implements low level logic for manipulating of
rpm/deb packages and repositories. Packetary can be used
both for creating partial and full mirrors of upstream and MOS
repos.
Change-Id: I8f4ccc71a27a486e6b0042eed6ce2ca8098deb48
DocImpact
Closes-Bug: #1517552
Previously, run_tests.sh was hardcoded to run fuel-web tests on py26.
Allow to pass the correct env list via $TOXENV environment variable.
Default to "py26" for backwards compatibility.
Needed to run py27 jobs on CI.
Change-Id: I3cfe5622bd0338f9a9a41ec6b6f6ac2d3c522009
Intern-based test suites are quite complete and stable and
should be used instead of outdated CasperJS-based suites.
Related to blueprint ui-functional-tests-with-intern
Partial-Bug: #1495536
Change-Id: I7272eb81fa54d914915ef45a0be777afa4b9d793
Can be used to skip UI compression to speed up tests development
Also fixed a bug with swallowing `gulp build` error
Change-Id: I914771bda53017e052a8432ac0180d9c91fb145b
There should be a way to use and test extensions when
there is no core database schema.
* perform data migration from the core only when there
is buffer table
* add basic tests which are run separately from core's
tests
Change-Id: Ide0d2e619913675ff4fff3febc9707bec3bfb112
Closes-bug: #1477020
Looks like we don't need it any more.
Besides there is a separate repository
https://github.com/stackforge/fuel-tasklib
Change-Id: I25b7f6d4ab9d6e3ff1d77d262913f4697557528e
We've removed fuel_agent into a separate repository,
so we don't need this directory here any more.
Closes-Bug: #1471849
Change-Id: I165259020721a4f6a87dce3a419bfdf64a5d0fcb
This commit makes it possible to execute UI
functional tests either based on Casper or on Intern/Selenium
Change-Id: I21d342093f142a1bbe7b8ae4f2c96fad1c125a0b
This commit adds unit-tests runner for UI and 1 test suite for
the expression parser. It also makes --webui flag for run_tests.sh
run all the UI tests: lint checks, unit tests, functional tests.
Change-Id: I6176dc01cdd6d2bcf7a99185e92cc498c7b2b899
New utility packages from a specified repo and
optionally configures it for a specified
Fuel environment.
Added fuel_package_updates to flake8 tests
blueprint package-fuel-components
Change-Id: I2d6d1f0d3823fe53b995d710ebce190ac1dab0b5
Also bump some libraries to the latest versions
Implements blueprint move-to-gulp
Related to blueprint ui-plugins
Change-Id: Iffcd5c2371a01969746ce4d18a88181246d674fd
The development started a year ago, since then
there are no updates of plans. The code is not
finished and outdated, hence it's just removing
of unused code.
Change-Id: I385d12f6161a03b43ab820821141f07e6b1365b9
Related-bug: #1434106
After python-fuelclient was moved to its own
repository the fuelclient folder can be deleted
from fuel-web.
New python-fuelclient's repository can be found here:
https://github.com/stackforge/python-fuelclient
Blueprint: re-thinking-fuel-client
Change-Id: Ic579bfdaacbafefc4050a6ec489888ced33d1b37
A new option -x added
run_tests.sh [-x | --performance]
- run nailgun performance tests
If -x option is not specified, run_test.sh will not run
performance tests. For this reason no matching -X option was added.
Option -n | -N works exactly as before: run | do not run
Nailgun non-performance tests.
Examples of usage:
run_tests.sh -n -x - run both regular and performance nailgun tests.
run_tests.sh -x - run nailgun performance tests only, do not run
regular nailgun tests
run_tests.sh -n - run regular nailgun tests only
run_tests.sh -N - run all tests except for nailgun regular and
performance tests
DocImpact
Change-Id: Id155e722da88549499641f5b924f35bfab3ab137
Closes-Bug: #1394196
Added verification that test file or directory exists for the command:
run_tests.sh -t <test file>
If a test file (or directory) does not exist, run_tests.sh prints error message:
ERROR: File or directory <test file> not found
and exits with status 1
Change-Id: Ia0b11efe434daf7f040943ea8622bad496074563
Closes-Bug: #1397916
Some tests require Nailgun to be running. For that a server is
started and is waited for 5 seconds. For some systems Nailgun
requires more time to get up and running.
This patch allows users to specify maximum time of waiting Nailgun
to start by setting NAILGUN_START_WAIT_TIME environment variable which
defaults to 5 seconds. Introduced hooks stop tests if Nailgun failed to
start before maximum wait-time passes.
DocImpact
Closes-bug: 1393566
Change-Id: I39d4b312510c34c190ad680c48f1e5792858b140
We need to flush database after running tests.
Before py.test with parallel tests was introduced
dropdb was performed by ./run_tests.sh, but currently
we need to run dropdb for each slave
- Added py.test --cleandb
- Connection should be closed and returned to pool,
and pool should be recreated, because we are reusing
same engine after db cleanup is done
- We should not reuse metadata used for models declaration,
it gets populated by stale db information
DocImpact
Closes-Bug: 1396500
Change-Id: Idc2d4d6395b8b5f30dc7f5e72a80b830808d2b75
* Options sorted alphabetically to find them more easily.
* Duplicated -l option changed to -k for tasklib tests.
Change-Id: I75056a593ee46f302c9916cc1201131965c1f657
Patch adds ability to execute test run distributed over
several processes, each one with isolated database.
Assumptions:
- nailgun user, requires createdb permission, if not please precreate
databases in format described below
- postgres database will be used for initial connection
- for environment where createdb can not be granted, like ci
operator need to create several databases, nailgun[0-7],
count of databases should be equal to TEST_WORKERS variable
- If no TEST_WORKERS provided - default database name will be used,
it is almost every time nailgun, but you can overwrite it with
TEST_NAILGUN_DB env variable
To execute tests on your local environment:
py.test -n 4 nailgun/tests/
DocImpact
Partial-Bug: 1388751
Change-Id: I6037108f1888b61f971a10db5209b20712467173
- List all plugins installed
fuel plugins
fuel plugins --list
- Install plugin
fuel plugins --install /tmp/sample.pb
1. Extracts metadata from plugin tar
2. If plugin successfully registered in nailgun
3. Untar all the files in /var/www/nailgun/plugins
- If you want to update plugin when it is already installed
add --force flag
fuel plugins --install /tmp/sample --force
The procedure is the same as for simple install, but
if plugin already registered - it will try to update
(send PUT request)
- Refactored tests to allow writing unit tests, with mock
and all that goodness
related to blueprint cinder-neutron-plugins-in-fuel
Change-Id: I642a300abcb9c50ef5039a343064f2ac78e39cd8
Medium between configuration providers and fuel orchestration
solution.
check README.md and functional tests for better understanding
how it works
HOW TO RUN:
python setup.py develop
taskcmd -c tasklib/tests/functional/conf.yaml run puppet/sleep
taskcmd -c tasklib/tests/functional/conf.yaml run exec/fail
taskcmd -c tasklib/tests/functional/conf.yaml daemon exec/long
taskcmd -c tasklib/tests/functional/conf.yaml status exec/long
Related to blueprint granular-deployment-based-on-tasks
Change-Id: Ifed1a9b90042bbbc93215a30dfb34e29dbe8fba2
Help scripts for deploy develpment environment on masternode:
- command line arguments handled
- deploy/revert for nailgun project
- requirements.txt installation into virtual env handled
- common functons for execute commands in containers
Change-Id: I3c60708e01867cd8455f79b3c02cfd505118cad5
Closes-Bug: #1279303
Andreas Jaeger