Fix LRT Azure failures: scope skip-list update to install/upgrade only

[Copilot]

LRT on Azure was failing due to skip-delete-resources-list.txt being unconditionally overwritten every run regardless of whether install/upgrade occurred.

Changes

.github/scripts/manage-radius-installation.sh

• Extract kubectl UCP resource snapshot into save_skip_resources_list() helper (with error handling); call it only in the install and upgrade branches — not the version-match no-op path

.github/workflows/long-running-azure.yaml

• Remove the unconditional "Create a list of resources not to be deleted" step; the snapshot now lives inside manage-radius-installation.sh and is scoped to install/upgrade only

Type of change

• This pull request fixes a bug in Radius and has an approved issue (issue link required).

Contributor checklist

• An overview of proposed schema changes is included in a linked GitHub issue.
  • Yes
  • Not applicable
• A design document PR is created in the design-notes repository, if new APIs are being introduced.
  • Yes
  • Not applicable
• The design document has been reviewed and approved by Radius maintainers/approvers.
  • Yes
  • Not applicable
• A PR for the samples repository is created, if existing samples are affected by the changes in this PR.
  • Yes
  • Not applicable
• A PR for the documentation repository is created, if the changes in this PR affect the documentation or any user facing updates are made.
  • Yes
  • Not applicable
• A PR for the recipes repository is created, if existing recipes are affected by the changes in this PR.
  • Yes
  • Not applicable

[lakshmimsft]

is the save needed each time workflow runs with a new key? perhaps the script can return if install/upgrade was done and we use that?

[codecov]

Codecov Report

✅ All modified and coverable lines are covered by tests.
✅ Project coverage is 51.23%. Comparing base (1506c2f) to head (ee0d860).

Additional details and impacted files

@@            Coverage Diff             @@
##             main   #11456      +/-   ##
==========================================
+ Coverage   51.21%   51.23%   +0.02%     
==========================================
  Files         699      699              
  Lines       44050    44050              
==========================================
+ Hits        22560    22571      +11     
+ Misses      19330    19326       -4     
+ Partials     2160     2153       -7     

☔ View full report in Codecov by Sentry.
📢 Have feedback on the report? Share it here.

🚀 New features to boost your workflow:

• ❄️ Test Analytics: Detect flaky tests, report on failures, and find test suite problems.
• 📦 JS Bundle Analysis: Save yourself from yourself by tracking and limiting bundle sizes in JS merges.

[github-actions]

Unit Tests

    2 files  ±0    415 suites  ±0   6m 23s ⏱️ -21s
4 862 tests ±0  4 860 ✅ ±0  2 💤 ±0  0 ❌ ±0 
5 764 runs  ±0  5 762 ✅ ±0  2 💤 ±0  0 ❌ ±0 

Results for commit ee0d860. ± Comparison against base commit 1506c2f.

♻️ This comment has been updated with latest results.

[Copilot]

Pull request overview

Fixes Azure long-running test (LRT) failures by ensuring skip-delete-resources-list.txt is only refreshed when Radius is actually installed/upgraded, instead of being overwritten on every run.

Changes:

• Move UCP resource snapshot generation into save_skip_resources_list() inside manage-radius-installation.sh.
• Call the snapshot helper only on install/upgrade paths (not on the version-match no-op path).
• Remove the unconditional workflow step that regenerated the skip list every run.

Reviewed changes

Copilot reviewed 2 out of 2 changed files in this pull request and generated 1 comment.

File	Description
.github/workflows/long-running-azure.yaml	Removes the always-run kubectl snapshot step; relies on the script to generate the skip list only when needed.
.github/scripts/manage-radius-installation.sh	Adds save_skip_resources_list() helper and invokes it from install/upgrade branches.

Comments suppressed due to low confidence (1)

.github/workflows/long-running-azure.yaml:282

• After removing the step that always generated skip-delete-resources-list.txt, this cache save step can run when the file was neither restored (cache miss) nor created (version-match no-op). actions/cache/save typically fails when the specified path doesn't exist, which would break the workflow even though skipping install/upgrade is a valid outcome. Guard this step with an if that checks the file exists (or create a minimal file in the no-op path without overwriting an existing one).

      - name: Save list of resources not to be deleted
        uses: actions/cache/save@668228422ae6a00e4ad889ee87cd7109ec5666a7 # v5.0.4
        with:
          path: skip-delete-resources-list.txt
          key: skip-delete-resources-list-${{ steps.gen-id.outputs.UNIQUE_ID || github.run_id }}

[Copilot]

save_skip_resources_list redirects output directly to skip-delete-resources-list.txt, which truncates the existing file before kubectl succeeds. If kubectl get fails (transient API error, RBAC, etc.), the workflow can lose the previously-restored skip list and later cleanup may delete resources that should have been preserved. Write to a temporary file and atomically replace the destination only on success (and leave the previous file intact on failure).

Suggested change

	if ! kubectl get resources.ucp.dev -n radius-system --no-headers -o custom-columns=":metadata.name" > skip-delete-resources-list.txt; then
	echo "Error: Failed to retrieve UCP resources from cluster." >&2
	exit 1
	fi
	echo "Skip resources list saved."
	local tmp_file
	tmp_file="$(mktemp skip-delete-resources-list.txt.XXXXXX)"
	if kubectl get resources.ucp.dev -n radius-system --no-headers \
	-o custom-columns=":metadata.name" > "${tmp_file}"; then
	mv "${tmp_file}" skip-delete-resources-list.txt
	echo "Skip resources list saved."
	else
	echo "Error: Failed to retrieve UCP resources from cluster." >&2
	rm -f "${tmp_file}"
	exit 1
	fi