...
- Updated suggestions from gerrit reviews https://gerrit.onap.org/r/79583:
Brian Freeman
if the intent is to have a aai-backwards-compatibility-v14.robot function that would get added to each release as the version number increases then I guess you need to create aai-[something]-v14, aai-something]-v16 etc. Not sure these are end to end tests becuase we arent testing that a Dublin vFW works end to end with v14 API's when most components will be calling the v16 APIs. If its simply AAI regression as part of verify/merge CSIT jobs then we should have a nomenclature with CSIT in the filename to distribution End to End tests from CSIT tests. aai-csit-v14 , aai-csit-v11 ?
Technically even Amsterdam had a mixture of versions supported so it seems like you may want to be testing by version not by release package name ?Daniel Rose
i looked over catherines presentation. I think if we are going to support csit type tests, i am ok. i think we need to put them in testsuites/<app> and tag those tests as csit. I still also dont think we should put release specific code in here though. I think supporting 2 versions of aai schema is ok if you ar releasing that, but calling something as casablanca is not - the branch should always work with the release it is.
Keong Lim
AAI team did have an idea that our tests could be used to diagnose installation problems, e.g. pulling the wrong version and wondering why it doesn't work. So as part of the "sanity integration" testing, we would publish these tools/scripts for use on any installation (development, test labs, production, etc), not just the latest.I will update a new patchset.
The naming can be updated. Wasn't sure if the release names were more familiar than the API version numbers.
Setup Procedure
Follow testsuite Readme for setup: https://gerrit.onap.org/r/gitweb?p=testsuite.git;a=blob;f=README.md;hb=HEAD
...
This section appears to define sub-routines that can be called by other "keywords". The "[Documentation]" appears to be a comment for the sub-routine. The "[Arguments]" appears to be a list of named parameters passed into the sub-routine. The lines that follow appear to be calls to other sub-routines in sequential order. Values can be returned from sub-routines and internal structures of the values can also be extracted and processed.
Worked Examples
/ https://gerrit.onap.org/r/79583Jira Legacy server System Jira serverId 4733707d-2057-3a0f-ae5e-4fd8aff50176 key AAI-2208
Running Robot
- In demo repository, "boot/robot/ete.sh" executes the "runTags.sh" script inside the openecompete_container
- In oom repository, "kubernetes/robot/ete-k8s.sh" executes the "runTags.sh" script inside robot pod
- In optf repository, "cmso/cmso-robot/docker/cmso-service/ete_test.sh" executes "docker-compose" to run the robot container
- In optf repository, "cmso/cmso-robot/ete.sh" executes the ROBOT_CMD="python -m robot.run" using the older method for Python 2.7 compatibility
- In testsuite repository, "runEteTag.sh" executes "python -m robot.run" connecting to an X-Windows display created by Xvfb and captures the command line output into a file
The robot command line needs:
- Path to robot library modules
- Path to robot variables file
- Tags of the tests to run
- A log-level setting
- An output file
- An input directory
See also "robot --help" or http://robotframework.org/robotframework/latest/RobotFrameworkUserGuide.html#id882
For example
Code Block | ||
---|---|---|
| ||
# Using Python 2.7 run-time
python -m robot.run \
--loglevel ${LOG_LEVEL} # aka -L option \
--outputdir ${OUTPUT_DIR} # aka -d option \
--variablefile ${VARIABLE_FILE} # aka -V option \
--pythonpath ${PATH_DIRECTORIES} # aka -P option \
--include ${TAGS_INCLUDED} # aka -i option \
--exclude ${TAGS_EXCLUDED} # aka -e option \
${DATA_SOURCES_DIRECTORY} |
robot will create an output XML file, a log HTML file and a report HTML file by default.
robot can also create documentation using the "python -m robot.libdoc" command.