search

giantswarm / inago

Inago orchestrates groups of units on fleet clusters
https://giantswarm.io/
Apache License 2.0
18 stars 2 forks source link

adds an int test that updates a unit with only one slice #233

Closed Nesurion closed 8 years ago

Nesurion commented 8 years ago

we don't have a test for this case yet.

codecov-io commented 8 years ago

Current coverage is 58.99%

Merging #233 into master will not affect coverage as of 367b7af

Powered by Codecov. Updated on successful CI builds.

Nesurion commented 8 years ago

@giantswarm/cena RFR

Nesurion commented 8 years ago

@xh3b4sd could you take another look?

xh3b4sd commented 8 years ago

I like that. Would you mind tuning the other tests as well?

Nesurion commented 8 years ago

Can do, but I'd like to get @zeisss and @JosephSalisbury opinion first. Maybe they remember why we put in the sleep in the first place.

JosephSalisbury commented 8 years ago

If i remember correctly, we had the sleep statements to give the system a beat to get consistent. If you get consistent good runs without them, we don't need them.

Nesurion commented 8 years ago

oh yeah I remember. Running the int-tests without the sleep resultet in random failing tests, which was super annoying. I'll add them back in.

xh3b4sd commented 8 years ago

But wait. Why is that then? Shouldn't we understand and find out the root cause?

JosephSalisbury commented 8 years ago

I think it might boil down to 'lol fleet', but looking into it further would be a good bug hunting expedition.

Nesurion commented 8 years ago

@giantswarm/cena RFR. I'll test things with lower or no sleep times in separate PR.

xh3b4sd commented 8 years ago

Was this actually related to a known bug we fixed? I would maybe point that out in the cram description. Besides that, LGTM.

Nesurion commented 8 years ago

I thought it was a bug, when it was just my local fleet being broken. Nevertheless I wanted to have a test that tests the update command against a single slice.