Race in waiting loop

[ericmillin]

I saw you reverted the resourceVersion solution. I have a fix for the race, though it does not do it "the right way" by watching resourceVersion. I can submit it if you think it is helpful as a temporary fix. There are a couple small change I will make before submitting, such as removing the paincs.

ericmillin@a523953

@rhuss

[dsimansk]

The fix was reverted as it seemed to cause instability in nightly/release CI jobs. However, it isn't manifesting now in the PR #1333. Could you please take a look, maybe we can iron out any possible nits?

[ericmillin]

@dsimansk Assuming that the wait will see the DELETED event even if it starts after deleteService is called, then it looks good to me.

[rhuss]

@ericmillin thanks for the heads up ! Let's close this issue and continue on #1333