Ultimate Mock Exam 2 : Help with question

I am stuck on a question of the Ultimate Mock Exam 2 (question 11).


The solution is to correct the livenessProbe port to 6443, which I did. Then wait for all the pods in kube-system to be running. In the second attempt, I also used “–watch” and waited for 3 minutes, since I was ahead of time. Everything remained in running state, yet in any attempt the results show the answer as wrong.

It’s hard to believe I am missing something, since the change is a simple one. So the question is, what am I missing? I also double checked the manifest after completing the mock.

manifest

Thanks :slight_smile:

As simple as the fix is, that is the fix: an improperly configured livenessProbe will bring down the apiserver

Thanks Rob. I should have been clear in my question. The main reason I am asking about this is that the result keeps coming up as done wrong. But at least I know I did not miss anything.

I’m not sure what’s up with the grader in those cases. I tried the problem today, and the grader was happy with the answer. Possibly YMMV

That’s OK. Thanks for the reply :slight_smile:

Hi @rob_kodekloud and @DanF I also I’m having this issue, sometimes the grader sent you a wrong answer, but in other times does approve your result. I’m not sure if checking the cluster just before ending the exam solves the issue temporally (Hope this could help you Danf if that’s an solution to know your exact grade). Also this question is repeated in other exams, I don’t know if in others exams just work well and there is the issue is just with the second one.

I plan to do Mock 3 today, and am still to do Mock 4. I’m done with this particular one as I only got 2 incorrect (including this one). Maybe it’s an issue in this particular one, but it’s not a big deal for me. If it is, it would be nice to be corrected for those who are still to do them, of course.

If this same question is in 3, 4, or 5… then I’ll post the results here. I had similar ones in the others, unsure if it was an Ultimate Mock, or a normal Mock that’s part of the CKA course… but in those cases it wasn’t the liveness port.

hi @DanF check this, it’s seem it’s solved:

You will have to wait the 15 second to have the kubectl access and then replace the kube-apiserver with the replace command and --force flag. That works for me now.

In my case I waited minutes, and everything was still running, which I why I can’t understand the wrong answer. Also I didn’t replace the kube-api, as it was running for a couple minutes without restarting. I also can’t justify why you’d need to replace it when these components are static pods. Am I right?

Appreciate the feedback :slight_smile: Also noticed you got a 100%… never managed to get to a 100% on these :woozy_face: well done :slight_smile:

1 Like

Yes, I also think that you shouldn’t need to force the restart, but I’m not so sure. What do you think @rob_kodekloud ?

Also @DanF the etcd-backup is now available, and in a couple of weeks you can make it the 100% if you are doing it everyday :slight_smile: