[BUG] Random behavior with pull-secret

[adrianriobo]

Tested on pre-release 1.31.2:

Pre conditions:

• fresh installation
• no pull secret set from config

Flow:

1. Start the cluster from the tray
2. When asking for pull-secret content paste something wrong (Ex. fdjdkgdjk)
3. Notification shows content is stored.
4. Notification about error with pull secret
5. Start again and ask for pull secret

Issues:

• Got a notification saying pull secret is stored when the validation prevents that.
• Random executions end up not picking the right pull-secret on the second time. I try to reproduce it but it is not consistent.
• On restart it asks again for pull secret

[adrianriobo]

Just comment I am hitting the pull secret issue lets say on 10% of starts.

[gbraad]

2. When asking for pull-secret content paste something wrong (Ex. fdjdkgdjk)
3. Notification shows content is stored.

We are now using the result of the POST of the pull-secret set command; 200 true, 500 = false

[gbraad]

Random executions end up not picking the right pull-secret on the second time. I try to reproduce it but it is not consistent.

TaskCancelled needs #130. Moved to a separate issue.

[gbraad]

This should have been resolved as tasks around the pullsecret will be retried when timeout and taskcanceled.

[adrianriobo]

did not hit again during testing on 1.33.1

[adrianriobo]

^^ @gbraad unfortunately I just hit again with version 1.33.1 after been testing it multiple times...lets say on 20 cluster starts / stop / start I hit it once

[gbraad]

What do you see when it happens? We forced an error.

Can you check the notifications?

[adrianriobo]

Did not see any notification, after stopping the cluster I just try to start again clicking on start menu item and the pull secret windows shows again (the pull secret was already used as the cluster was already started).

This was ran from ux testing automation so the test end up failing as on the restart this is not expected:

<testsuite name="UX Test" tests="6" skipped="0" failures="3" errors="0" time="1621.6018907">
    <testcase name="Install tray" status="passed" time="1.1290731"></testcase>
    <testcase name="Start the cluster" status="passed" time="580.1197926"></testcase>
    <testcase name="Connect to the cluster #1" status="failed" time="14.8209602">
      <failure message="Step user is connected to the cluster as kubeadmin: output did not match. Expected: &#39;kubeadmin&#39;, Actual: &#39;&#39;"></failure>
    </testcase>
    <testcase name="Connect to the cluster #2" status="failed" time="15.2240385">
      <failure message="Step user is connected to the cluster as developer: output did not match. Expected: &#39;developer&#39;, Actual: &#39;&#39;"></failure>
    </testcase>
    <testcase name="Stop the cluster" status="passed" time="84.9260149"></testcase>
    <testcase name="Restart the cluster" status="failed" time="925.3369163">
      <failure message="Step cluster should be running: not found: running. Timeout"></failure>
      <error message="Step tray should show cluster as running" type="skipped"></error>
      <error message="Step user should get notified with cluster state as running" type="skipped"></error>
    </testcase>
  </testsuite>