test: various fixes to package testing

[patrick-stephens]

Resolves various test failures seen during recent builds and releases.

• Version value was not being passed through correctly in some cases
• Install script can fail if index page is still working on providing versions
• Too much debug from installing helper things on some platforms
• Curl hitting timeouts as well for accessing the index page

---

Summary by cubic

Stabilizes package and container tests by fixing version env propagation and installer retries, and corrects tag handling for PR builds. Also generates JSON schema artifacts from container images during build (non-blocking).

• Bug Fixes

  • Install script now retries version index fetches (up to 5 attempts with 10s backoff) and ignores CDN “just a moment” pages before failing.
  • Use a 60s curl timeout for index checks in the installer and integration tests.
  • Propagate version env correctly: set TELEMETRY_FORGE_AGENT_VERSION (and legacy FLUENTDO_AGENT_VERSION) from the image tag in the workflow, default version to tag in test runners, and fix tag usage for PR builds.

• Test Updates

  • Reduce noise in container installs (yum -yq, apt-get -qqy).
  • Pass TELEMETRY_FORGE_AGENT_VERSION and TELEMETRY_FORGE_AGENT_URL into functional test containers.
  • Clarify Bats test names to include version/binary for easier debugging.

^{Written for commit 89f3891. Summary will update on new commits.}

[cubic-dev-ai]

2 issues found across 7 files

Confidence score: 3/5

• There is a concrete reliability risk in install.sh: without curl timeouts, the current retry loop can still block indefinitely before retries, which can leave installs hanging for users.
• In testing/bats/tests/integration/linux/integration-install-script.bats, using --connect-timeout alone does not cover slow response bodies, so tests may miss the same hang condition once a connection is established.
• This is a moderate, user-impacting timeout behavior issue (with high confidence), so merge risk is not low until total request timeouts are applied consistently.
• Pay close attention to install.sh and testing/bats/tests/integration/linux/integration-install-script.bats - timeout handling should cover the full request lifecycle, not just connection setup.

Prompt for AI agents (unresolved issues)

Check if these issues are valid — if so, understand the root cause of each and fix them. If appropriate, use sub-agents to investigate and fix each issue separately.


<file name="testing/bats/tests/integration/linux/integration-install-script.bats">

<violation number="1" location="testing/bats/tests/integration/linux/integration-install-script.bats:38">
P2: `--connect-timeout` only covers TCP connection setup here; it will not help when `/index.html` is slow to return after the connection is established. Use a total request timeout instead.</violation>
</file>

<file name="install.sh">

<violation number="1" location="install.sh:378">
P2: Add curl timeouts here; otherwise this retry loop can still hang forever before any retry happens.</violation>
</file>

_{Reply with feedback, questions, or to request a fix. Tag @cubic-dev-ai to re-run a review.}

[patrick-stephens]

@cubic-dev-ai please re-review

[cubic-dev-ai]

@cubic-dev-ai please re-review

@patrick-stephens I have started the AI code review. It will take a few minutes to complete.

[cubic-dev-ai]

No issues found across 7 files

Confidence score: 5/5

• Automated review surfaced no issues in the provided summaries.
• No files require special attention.

[cubic-dev-ai]

1 issue found across 3 files (changes from recent commits).

Prompt for AI agents (unresolved issues)

Check if these issues are valid — if so, understand the root cause of each and fix them. If appropriate, use sub-agents to investigate and fix each issue separately.


<file name=".github/workflows/call-test-containers.yaml">

<violation number="1" location=".github/workflows/call-test-containers.yaml:90">
P2: The tag-only `if` condition prevents setting version env vars for non-tag runs, so container tests can validate against the wrong fallback version instead of `inputs.image-tag`.</violation>
</file>

_{Reply with feedback, questions, or to request a fix. Tag @cubic-dev-ai to re-run a review.}

[cubic-dev-ai]

P2: The tag-only if condition prevents setting version env vars for non-tag runs, so container tests can validate against the wrong fallback version instead of inputs.image-tag.

Prompt for AI agents

Check if this issue is valid — if so, understand the root cause and fix it. At .github/workflows/call-test-containers.yaml, line 90:

<comment>The tag-only `if` condition prevents setting version env vars for non-tag runs, so container tests can validate against the wrong fallback version instead of `inputs.image-tag`.</comment>

<file context>
@@ -88,6 +86,13 @@ jobs:
         shell: bash
 
+      - name: Ensure we use the right version during a release
+        if: startsWith(github.ref, 'refs/tags/v')
+        run: |
+          echo "TELEMETRY_FORGE_AGENT_VERSION=${{ inputs.image-tag }}" >> $GITHUB_ENV
</file context>

Suggested change

	if: startsWith(github.ref, 'refs/tags/v')
	if: ${{ always() }}