-
Notifications
You must be signed in to change notification settings - Fork 4.3k
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
Flaky: apache_beam.runners.portability.fn_api_runner.fn_runner_test.FnApiRunnerTestWithGrpcAndMultiWorkers #21104
Comments
There have been several fixes (see Jira). last one: https://github.com/apache/beam/pull/17537/files However this test is still flaky. Seeing a different error this time in apache_beam.runners.portability.fn_api_runner.fn_runner_test.FnApiRunnerTestWithGrpcAndMultiWorkers.test_pardo_large_input (from py38-cython) :
|
I'm having lots of trouble reproducing this failure... |
but I'm trying to : ) |
may be one of those tests that fail on jenkins only (when run together w/ a bunch of tests)... i saw it today in a linked issue |
i got it to reproduce a couple of times on my machine in non-cython. I'm adding logging to see what happens |
nice, did you run a single test? |
I got it to repro by runnig the full class of tests (FnApiRunnerTestWithGrpcAndMultiWorkers) |
fwiw this is the error stack that I'm able to reproduce:
|
Errors I'm seeing:
And afterwards
|
I've done some more research. Oddly, the encoding size for the first set of elements that we output varies even on instances of successful tests. I wondered if this could be because we encode every element with the current timestamp, which varies depending on test execution time - so I hardcoded a specific timestamp for all instances of the test. Encoding size still varied:
|
@pabloem are you still actively working on this? |
Thanks Kenn. I am not working actively on this now.
If I find some time I'll get back to it - but for now anyone else should
feel free to pick it up.
…On Wed, Dec 7, 2022, 8:31 AM Kenn Knowles ***@***.***> wrote:
@pabloem <https://github.com/pabloem> are you still actively working on
this?
—
Reply to this email directly, view it on GitHub
<#21104 (comment)>, or
unsubscribe
<https://github.com/notifications/unsubscribe-auth/AAJ5Z3DN5XMZWS2F2HVFYXLWMC3WXANCNFSM5X4AZMYA>
.
You are receiving this because you were mentioned.Message ID:
***@***.***>
|
I'll unassign you then |
I think this is fixed. If not, it should get auto-flagged by our tooling anyways, so this should be safe to close |
The
apache_beam.runners.portability.fn_api_runner.fn_runner_test.FnApiRunnerTestWithGrpcAndMultiWorkers
tests are flaky and causing precommit failures that seem similar.test_pardo_windowed_side_inputs
: https://ci-beam.apache.org/job/beam_PreCommit_Python_Cron/4417/consoletest_pack_combiners
: https://ci-beam.apache.org/job/beam_PreCommit_Python_Cron/4415/consoleFullImported from Jira BEAM-12603. Original Jira may contain additional context.
Reported by: tysonjh.
The text was updated successfully, but these errors were encountered: