Fix flaky write_with_sse_kms_key_id_ok test #1140
Conversation
Signed-off-by: Christian Hagemeier <chagem@amazon.com>
c-hagem
temporarily deployed
to
PR integration tests
GitHub Actions
Inactive
c-hagem
temporarily deployed
to
PR integration tests
GitHub Actions
Inactive
c-hagem
temporarily deployed
to
PR integration tests
GitHub Actions
Inactive
c-hagem
temporarily deployed
to
PR integration tests
GitHub Actions
Inactive
c-hagem
temporarily deployed
to
PR integration tests
GitHub Actions
Inactive
c-hagem
temporarily deployed
to
PR integration tests
GitHub Actions
Inactive
c-hagem
temporarily deployed
to
PR integration tests
GitHub Actions
Inactive
| let f_name = "f.txt"; | ||
| write_to_file(mount_point.path(), f_name).expect("should be able to write to the file"); | ||
| unmount(&mount_point); | ||
| wait_for_exit(child); |
There was a problem hiding this comment.
Is this the right example of an error which you're intending to fix?
In which case, I am not sure if I understood the fix. The test fails before wait_for_exit: in write_to_file, called on line 833. Do you mind explaining the fix in more detail? (it's a good practice to add a comment in code with such explanation)
There was a problem hiding this comment.
I'm not sure if we should add a comment since it looks like we're doing the right thing everywhere else. However, I agree that the fix here is currently unclear and we should make sure that the PR description (and later this will be the commit) includes why this was a problem.
There was a problem hiding this comment.
Agree, no comments needed and if it helps to resolve issues locally, it would be great to merge this and see if it helps in the CI. Just curious about the idea behind the fix.
There was a problem hiding this comment.
I've set this to merge for now since its bringing this test consistent with the other tests. I'm still surprised that this would fix the issue.
github-merge-queue
bot
removed this pull request from the merge queue due to failed status checks
Fixes the flakiness of write_with_sse_kms_key_id_ok test, which was caused by not properly unmounting and dropping child.
This is not a breaking change; no changelog entry required (as this just fixes a test).
Before this change, this test fails in ~10 out of 100 runs, after this change it fails 0 times out of 100 runs.
By submitting this pull request, I confirm that my contribution is made under the terms of the Apache 2.0 license and I agree to the terms of the Developer Certificate of Origin (DCO).