pgj commented on code in PR #5341:
URL: https://github.com/apache/couchdb/pull/5341#discussion_r1863840544
##########
Makefile:
##########
@@ -339,7 +339,7 @@ mango-test: devclean all
--admin=adm:pass \
--no-eval "\
COUCH_USER=adm COUCH_PASS=pass \
-src/mango/.venv/bin/nose2 -s src/mango/test -c src/mango/unittest.cfg
$(MANGO_TEST_OPTS)"
+src/mango/.venv/bin/nose2 -F -s src/mango/test -c src/mango/unittest.cfg
$(MANGO_TEST_OPTS)"
Review Comment:
> 0.85 sec vs 50 sec to run the entire test suite.
Sure, I know that running the Mango test suite in the past required around a
minute to run. In my opinion, that is acceptable, since it is still does not
take tens of minutes or many hours.
> Different combinations of test cases will result in the same failure
message but fail on different test cases.
I am not sure if I got you on this. If the "fail fast" mode stops at the
first failing test, how can it result in the failure message if there are
multiple failures?
> `-F, --failfast` allows us to identify the issue quickly, speed up the
feedback loop, and make debugging easier.
Note that you could tell the names of the modules or specific test cases on
the command line when invoking `nose2`. This is also exposed in the general
`Makefile` of the repository, for the `mango-test` target as the
`MANGO_TEST_OPTS` variable.
With the help of this, one could easily limit the current run to the tests
that one has to run for keeping the feedback loop short. With the regards to
finding the tests to run, it is okay to run the whole suite once and take the
cost of the full run.
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
To unsubscribe, e-mail: [email protected]
For queries about this service, please contact Infrastructure at:
[email protected]
