Adding test for extension group example #5
Conversation
This is have better separation between them and the python tests. BUCK files have been updated accordingly. Changes to the internal generateCopyFileTarget function were needed. The function now supports a base folder to be set so that the regex or file path is appended to that base path, without having it included in the destination. It will also not use a library target anymore, but a custom one so that's possible to set properties with custom names.
- Add the bdist_wheel package to have pip optimize packages. - Update Python 2 packages to Python 3 - Use pexpect==3.3
Restored all the python tests under tools/tests. Some tests have been modified so that they can be relocated and don't expect too specific hardcoded paths. Fixed the gen_api function in genapi.py missing the path to the spec file when evaluating blacklisted tables. Updated the wiki docs with the optionally needed packages to run the python tests. Updated azure-pipelines.yml to install the needed packages for the python tests. Add an init process so that reaping works in the CI Docker image
- Update to Python3. - Scripts now accepts the test configs directory as an argument. - Update findOsquerydBinary to support Ninja builds - Fix test_windows_service Administrator detection - Avoid a ResourceWarning on BufferedReader when killing the osquery process
- Fixed a ResourceWarning due to a child killed but not yet reaped.
Magic tables requires a 'where' clause: kolide/fleet#1040
The timeout concept was confusing, because it wasn't really waiting for that amount of time and then interrupting the tentatives. It was just trying timeout/interval times which by default, in the expectTrue case, was 800 times, which is a lot and compounds to a lot if the function tested did some tentatives too. timeout is now called attemps, and the default values have been changed to something a bit saner.
psutil caches the results, so when accessing a previously found process, it might hand a dead process handle, so accessing it throws an exception. Handle the exception and ignore the process.
- attempt variable has to be increased each loop, otherwise we end in an infinite loop. - Corrected example_extension.ext path.
| # There is a brief race between register and registry broadcast | ||
| # That fast external client fight when querying tables. | ||
| # Other config/logger plugins have wrappers to retry/wait. | ||
| time.sleep(0.5) |
There was a problem hiding this comment.
If this is caused by the fact that an extension is not immediately loaded when osquery starts, you can ask osquery to wait until such extension is loaded with --extensions_require=<extension name>.
Or it will wait at least up to the default interval (which is 3 seconds).
In this way you could also check that no error message is printed; specifically:
An error occured during extension manager startup: Required extension not found or not loaded: <extension name you used in the option>
There was a problem hiding this comment.
Cool. I just reused code from existing tests though :)
There was a problem hiding this comment.
Ah oups, didn't saw that.
Well ok, it's true that the test will still fail if the time it waits is not enough because the results will be empty, but you will have a less precise knowledge of why.
Anyway, it will be for another time then.
There was a problem hiding this comment.
I can fix both this and test_91_extensions_settings then. I mean now!
Smjert
force-pushed
the
stefano/test/python-tests
branch
from
10992bf
to
d17641a
Compare
|
Current status:
|
Thanks @kumarak @Smjert