Age | Commit message (Collapse) | Author | Files | Lines |
|
Change-Id: Id94475dd519ce04b87da470f6e6bb5a1c1aa2d75
Signed-off-by: Cédric Ollivier <cedric.ollivier@orange.com>
|
|
It simply converts run_tests.py's functions to Runner's methods.
This proposal could be enhanced in a second step to fully leverage on
object programming.
It defines as instance variables the former static variables to allow
multiple runs (it's mandatory for unit testing).
All false positives in unit tests are fixed but run_tests.py is still
not fully covered.
Change-Id: I8f91c132aa8b2248041f31e46238dd5598344d34
Signed-off-by: Cédric Ollivier <cedric.ollivier@orange.com>
|
|
This will allow enable/disable test cases directly
using true/false.
Change-Id: Iaa7d051ebb8dd9e5a6f835a823f58f1e17a87b8d
Signed-off-by: jose.lausuch <jose.lausuch@ericsson.com>
|
|
run_tests.py now relies on PrettyTable as most of the openstack clients.
generate_report.py and its related unit tests are simply removed.
It sets padding_width=5 in testcase.py too to conform with
run_tests.py.
Now report is printed in every case.
Change-Id: Id9ce93f984503f25d6a2150482f397853fa3dd64
Signed-off-by: Cédric Ollivier <cedric.ollivier@orange.com>
|
|
It takes into account the next framework modifications:
- Switch TestCase attribute criteria to result [1]
- Manage criteria in TestCase [2]
- Print the real testcase duration [3]
- Rename TestCase check_result() to is_successful() [4]
[1] https://gerrit.opnfv.org/gerrit/#/c/33695/
[2] https://gerrit.opnfv.org/gerrit/#/c/33697/
[3] https://gerrit.opnfv.org/gerrit/#/c/34369/
[4] https://gerrit.opnfv.org/gerrit/#/c/34373/
Change-Id: I3a6790e4233ef4c55b0535618dacf702f779f28b
Signed-off-by: Cédric Ollivier <cedric.ollivier@orange.com>
|
|
The tier name was incorrect when passing a test case as run_test.py arg.
Change-Id: I6f439fc7b90facb3f0553798bd1fe06ac9f09475
Signed-off-by: Cédric Ollivier <cedric.ollivier@orange.com>
|
|
|
|
Robotframework sshlibrary monkeypatches paramiko runtime and version
2.1.1 causes
2017-03-27 15:24:23,196 - SSH utils - ERROR - _custom_start_client() got
an unexpected keyword argument 'timeout'
errors in SFC test because of the issue
https://github.com/robotframework/SSHLibrary/issues/158
This issue is resolved in version 2.1.3
Change-Id: I07f570c1a90f358b9f17aa6c959642d67e07a7e4
Signed-off-by: George Paraskevopoulos <geopar@intracom-telecom.com>
|
|
|
|
It adds PrettyTable as requirement even if it's already defined in
OpenStack client dependencies.
If the TestCase object is considered as invalid, it simply returns the
default str.
Change-Id: Iee788aef2a13694d9482560977cbbf21c7f2c967
Signed-off-by: Cédric Ollivier <cedric.ollivier@orange.com>
|
|
|
|
* changes:
Remove the former results in DB url
Implement TestCase __str__()
|
|
|
|
It conforms generate_report.py with the previous change [1].
[1] https://gerrit.opnfv.org/gerrit/#/c/33511/
Change-Id: I677ebd416f462a591439fc7b637c02d7964e6173
Signed-off-by: Cédric Ollivier <cedric.ollivier@orange.com>
|
|
It adds helpers to send notifications to Energy recording API and
related unit tests.
It requires a dedicated section in functest config file
to set connectivity parameters to Energy recording API.
It is using shared API Recording at http://161.105.253.100:8888
Change-Id: Idcb74d1bf7341ccce7cc1c3926f22338ce24f714
Signed-off-by: Benoit HERARD <benoit.herard@orange.com>
|
|
Now ODL TestCase calculates the ratio between critical tests passed
and failed and saves it in result. Non-critical test cases can fail as
result doesn't take them into account [1].
It also updates default attribute values in TestCase and allows result
to be float.
[1] http://robotframework.org/robotframework/latest/RobotFrameworkUserGuide.html#setting-criticality
Change-Id: Id4a89271b5b1a90cd3c1e2b08591ff26ffaffee0
Signed-off-by: Cédric Ollivier <cedric.ollivier@orange.com>
|
|
Change-Id: I514340fc6180e29b31ef11bdea9f719aa2bbb738
Signed-off-by: Cédric Ollivier <cedric.ollivier@orange.com>
|
|
Change-Id: I0b8b9b2e85717fc92a233e4f7344b3419421778d
Signed-off-by: Cédric Ollivier <cedric.ollivier@orange.com>
|
|
Testcases about snaps failed due to the absence of attribute 'logger'.
Change-Id: I78c85e51f63afa767fa73aa12faea066ad432371
Signed-off-by: Linda Wang <wangwulin@huawei.com>
|
|
It's designed to be called by run_tests.py when printing the results of
test cases and when generating the global report.
Change-Id: If4a6f023ef2344bbc4f940d07dde4b776dce5d68
Signed-off-by: Cédric Ollivier <cedric.ollivier@orange.com>
|
|
|
|
|
|
Change-Id: I41cea33b4c1dae79e75231e7c7e069fc89736f34
Signed-off-by: Cédric Ollivier <cedric.ollivier@orange.com>
|
|
TestCase offers a new public method to calculate and print the duration.
It also adds the related unit tests and adapts run_tests.py.
Change-Id: Ib6a7e637dafacb9027146199aeb033c2dcb986c6
Signed-off-by: Cédric Ollivier <cedric.ollivier@orange.com>
|
|
|
|
|
|
Directories affected:
- ci
- core
and respective unit tests
Change-Id: I6a3d5aa68de29fc5a37ae543a067ff797eba33e6
Signed-off-by: jose.lausuch <jose.lausuch@ericsson.com>
|
|
|
|
|
|
|
|
It avoids erasing functest.log when project_name = functest.
Change-Id: Icae898abb8b51c89b79bb1124adfadab8b0b3b99
Signed-off-by: Cédric Ollivier <cedric.ollivier@orange.com>
|
|
Package iniparse is absent for functest, while it is required by
rally.py
Change-Id: Ia50216cfebdc9033f6a5f4abc6338951aaa2e8ab
Signed-off-by: Linda Wang <wangwulin@huawei.com>
|
|
Change-Id: I4b0060502621c7491820b9aa42d8f2f63b06a9ae
Signed-off-by: Cédric Ollivier <cedric.ollivier@orange.com>
|
|
Change-Id: Icc09b824a4e7906dfab031b679ba3d0aa86ab0cd
Signed-off-by: Morgan Richomme <morgan.richomme@orange.com>
|
|
|
|
|
|
|
|
Change-Id: Ic9ea7507dc5f1f781b57c4573e971a2c8a8ca9ef
Signed-off-by: Cédric Ollivier <cedric.ollivier@orange.com>
|
|
This mangling ensures that all info messages printed from core packages
are shown in console. It also avoids sphinx to print them.
Change-Id: I07db9f33060c195bce3b48b06a6640eb6c56c2eb
Signed-off-by: Cédric Ollivier <cedric.ollivier@orange.com>
|
|
It deletes functest.utils.functest_logger and the related unit tests.
Then it modifies all functest modules to get all loggers via
logging.getLogger(). __name__ is mainly used as getLogger arg as
proposed by logging [1].
All loggers and handlers are now defined via functest/ci/logging.ini
instead of a dict loaded by an external json file.
Now only warn messages and info messages from ci and core packages are
printed in console.
[1] https://docs.python.org/2/library/logging.html
Change-Id: Ic192855e0f9bf94825d8f7ec73549a0f3b8d44c5
Signed-off-by: Cédric Ollivier <cedric.ollivier@orange.com>
|
|
|
|
|
|
|
|
|
|
* changes:
Remove openstack directory
Remove create_instance_and_ip.py
|
|
|
|
Change-Id: Ifeb4aecf66d75c9bb98d5dafdbe20be78423565e
Signed-off-by: helenyao <yaohelan@huawei.com>
|
|
|
|
It is useles right now and won't be used in the future.
Change-Id: I9aa121544051f49f7e7e794ae3b2c4a9ea516e47
Signed-off-by: jose.lausuch <jose.lausuch@ericsson.com>
|
|
1. The endpoint type to use for OpenStack Identity API v3
"v3_endpoint_type" is set to be "adminURL" by default, which should
be updated to the env OS_ENDPOINT_TYPE with the value "publicURL"
2. Reset the identity region name to "RegionOne"
3. For keystone api v3 in Newton, 5 testcases aimming at identity admin
v2 test should be omitted.
JIRA: FUNCTEST-801
Change-Id: I47ae41a80de0a52d32d25c9783860955f465099c
Signed-off-by: Linda Wang <wangwulin@huawei.com>
|