test: fix unset variable in preconfig integration test

[annekebr]

As the actual update of the config values happens in its own bash function, the IPP variable is no longer set when performing the values.yaml update resulting in an empty value in the yaml file (imagePullPolicy: "").

Fixes #

Description

Checklist

• PR is rebased to/aimed at branch develop
• PR follows Contributing Guide
• Added tests (if necessary)
• Extended README/Documentation (if necessary)
• Adjusted versions of image and Helm chart in values.yaml and Chart.yaml (if necessary)

[codecov-commenter]

Codecov Report

Merging #542 (c85c4bb) into develop (29803af) will not change coverage.
The diff coverage is n/a.

@@           Coverage Diff            @@
##           develop     #542   +/-   ##
========================================
  Coverage    94.40%   94.40%           
========================================
  Files           22       22           
  Lines         1125     1125           
========================================
  Hits          1062     1062           
  Misses          63       63           

---

Continue to review full report at Codecov.

Legend - Click here to learn more
Δ = absolute <relative> (impact), ø = not affected, ? = missing data
Powered by Codecov. Last update 29803af...c85c4bb. Read the comment docs.

[xopham]

happy to approve this. however, does it actually cause an error?
if so, please update the description...

[annekebr]

@xopham I'm pretty sure that it was intended to actually replace the value for the imagePullPolicy in https://github.com/sse-secure-systems/connaisseur/blob/master/tests/integration/integration-test.sh#L286 by the value IPP that seems to be set only for this purpose, but as the actual update is invoked in its own function the variable is no longer set. I was wondering why the test still works as Kubernetes says that it only allows Never, IfNotPresent and Always as imagePullPolicy according to their docs, but in my local tests an empty string seems to be valid as well. Didn't allow that in #458 though, that's why I stumbled across it. I updated #458 accordingly, but I thought that this might be worth fixing anyway before it actually causes issues.

[xopham]

Thx for updating 🙏
Might that have caused #497 ?

[xopham]

Lgtm

[annekebr]

It seems that the issue was introduced in #497, but as imagePullPolicy="" defaults to Always it still worked as intended :)