Hive has >24hours of tests - in case of automated reruns... I wonder how a
patch which breaks almost all tests should be handled?
I believe we already have a process to step up against these things: if you
encounter a flaky test - it should be checked&disabled by using:
http://ci.hive.apache.or
I’ll second the above. We already have a number of flaky tests that would go
undetected and just increase the build time.
Hello,
I tend to agree with Sai; if we can run on demand the failed tests
that would be a cool feature. If we just rerun everything that fails
without questions asked we may do more harm than good. Taking into
account the fact that nobody runs all tests locally before submitting
a PR (cause it is
Hello everyone,
My personal preference is that the option to rerun *only* the failed test
suites should be manual. The reason is, there might be frequent intentional
test failures (failures due to our own patch) and the tests being flaky is
less likely. So giving the option to the user to rerun th
+1 from me as well to rerun the failing tests.
The oracle docker is also a pain, it is one of the main reasons for
retriggers, these retriggers wastes a lot of resources and increases the
time to get build results for genuine runs.
-Ayush
On Thu, 8 Jun 2023 at 12:31, Butao Zhang wrote:
> +1. I
+1. I often have to rerun whole pre-commit job due to individual unstable test,
ant it is too time-consuming. It would be much better if we can set maven
config to retry automatically.
Thanks,
Butao Zhang
Replied Message
| From | r12 t45 |
| Date | 6/8/2023 14:52 |
| To | |
| Subj
