fix(client): retry transient API failures

[jingxiang-z]

Description

Checklist

• I am familiar with the Contributing Guidelines.
• My commits are signed off (git commit -s) per the Contributing Guidelines.
• New or existing tests cover these changes.
• The documentation is up to date with these changes.

Summary by CodeRabbit

• New Features

  • Added automatic retries for transient network failures and selected temporary server errors.
  • Honors server-provided retry timing and stops promptly when requests are canceled.
  • Signing-key retrieval now benefits from the same retry behavior.

• Bug Fixes

  • Paginated “list all” operations retry only the failed page, avoiding duplicate requests and results.
  • Improved pagination error messages with page and progress details.

[coderabbitai]

No actionable comments were generated in the recent review. 🎉

ℹ️ Recent review info

⚙️ Run configuration

Configuration used: Path: .coderabbit.yaml

Review profile: CHILL

Plan: Enterprise

Run ID: dd50e7a8-3080-4df5-8cbc-ae78ad22b302

📥 Commits

Reviewing files that changed from the base of the PR and between 24e5508 and 85fee0c.

📒 Files selected for processing (3)

• cmd/nvfleetint/node_test.go
• nvfleetint/client.go
• nvfleetint/client_test.go

🚧 Files skipped from review as they are similar to previous changes (3)

• cmd/nvfleetint/node_test.go
• nvfleetint/client_test.go
• nvfleetint/client.go

---

📝 Walkthrough

Walkthrough

The client adds bounded retries for idempotent requests, including Retry-After handling and cancellation-aware backoff. Signing-key retrieval uses the retry path. Pagination reports contextual errors, and tests verify failed-page-only retries without duplicate nodes.

Changes

HTTP retry engine

Layer / File(s)	Summary
Bounded request retry engine nvfleetint/client.go, nvfleetint/client_test.go	The client retries selected transient GET and HEAD failures with bounded jittered backoff, Retry-After support, response cleanup, and context cancellation. Tests cover transient and permanent statuses, delay parsing, and cancellation.

Signing-key retry integration

Layer / File(s)	Summary
Signing-key retry integration nvfleetint/client.go, nvfleetint/verify_test.go	Signing-key retrieval uses the retrying request doer. Tests verify recovery from a transient 503 response.

Pagination validation

Layer / File(s)	Summary
Pagination retry and error reporting internal/clihelpers/pagination.go, cmd/nvfleetint/node_test.go	Pagination errors include the item, API page, and completed-page count. The node listing test verifies that only the failed page is retried and nodes are not duplicated.

Estimated code review effort: 4 (Complex) | ~45 minutes

Sequence Diagram(s)

sequenceDiagram
  participant Client
  participant RetryingRequestDoer
  participant HTTPServer
  Client->>RetryingRequestDoer: Send idempotent GET or HEAD
  RetryingRequestDoer->>HTTPServer: Execute request
  HTTPServer-->>RetryingRequestDoer: Return transient failure
  RetryingRequestDoer->>RetryingRequestDoer: Apply Retry-After or backoff
  RetryingRequestDoer->>HTTPServer: Retry request
  HTTPServer-->>Client: Return final response

Loading

🚥 Pre-merge checks | ✅ 5

✅ Passed checks (5 passed)

Check name	Status	Explanation
Description Check	✅ Passed	Check skipped - CodeRabbit’s high-level summary is enabled.
Title check	✅ Passed	The title clearly and concisely describes the main change: retrying transient API failures.
Docstring Coverage	✅ Passed	No functions found in the changed files to evaluate docstring coverage. Skipping docstring coverage check.
Linked Issues check	✅ Passed	Check skipped because no linked issues were found for this pull request.
Out of Scope Changes check	✅ Passed	Check skipped because no linked issues were found for this pull request.

✨ Finishing Touches

📝 Generate docstrings

• Create stacked PR
• Commit on current branch

🧪 Generate unit tests (beta)

• Create PR with unit tests
• Commit unit tests in branch fix/retry-transient-api-requests

---

_{Comment @coderabbitai help to get the list of available commands.}

[coderabbitai]

Actionable comments posted: 2

🤖 Prompt for all review comments with AI agents

Verify each finding against current code. Fix only still-valid issues, skip the
rest with a brief reason, keep changes minimal, and validate.

Inline comments:
In `@cmd/nvfleetint/node_test.go`:
- Around line 327-340: Strengthen the merged-output assertions in the test
around the decoded got.Items value: unmarshal each item’s nodeUUID and verify
the returned identities are node-1 followed by node-2. Keep the existing
item-count, pagination, and page-call assertions, but ensure the test rejects
duplicate first-page nodes.

In `@nvfleetint/client.go`:
- Around line 388-389: Update the Retry-After parsing logic around the seconds
conversion to validate that the parsed value does not exceed the maximum
representable time.Duration in seconds before multiplying by time.Second; reject
or clamp oversized values so the resulting duration never wraps negative. Add
boundary coverage for the largest valid value and an overflowing value,
including the resulting waitForRetry behavior.

🪄 Autofix

Fix all unresolved CodeRabbit comments on this PR:

• Push a commit to this branch (recommended)
• Create a new PR with the fixes

---

ℹ️ Review info

⚙️ Run configuration

Configuration used: Path: .coderabbit.yaml

Review profile: CHILL

Plan: Enterprise

Run ID: fed230a7-3229-4c71-8398-83ef883a54b8

📥 Commits

Reviewing files that changed from the base of the PR and between 0f5a58a and 3a0f8d7.

📒 Files selected for processing (5)

• cmd/nvfleetint/node_test.go
• internal/clihelpers/pagination.go
• nvfleetint/client.go
• nvfleetint/client_test.go
• nvfleetint/verify_test.go