Avoid expect_any_instance_of in transfer service spec
What does this MR do?
Avoid expect_any_instance_of
because it doesn't work well with prepend
Are there points in the code the reviewer needs to double check?
Sometimes it's extremely convenient with expect_any_instance_of
or allow_any_instance_of
, however it does introduce some issues with prepend
, as we're using it more and more.
It's very hard to rewrite all the existing tests, and it could also be much more complicated to do so.
In an ideal world, we should not use them, but it's certainly attractive to use them. I am not sure how to find the right balance here.
Why was this MR needed?
Fixes https://gitlab.com/gitlab-org/gitlab-ee/-/jobs/26183762 (Note that this only happens in EE because EE has prepend
while CE doesn't)
What are the relevant issue numbers?
This was discovered in https://gitlab.com/gitlab-org/gitlab-ee/merge_requests/2554#note_36312731