improving several tests setup and instructions

[Sheeproid]

Summary by CodeRabbit

• Tests
  • Enhanced test fixtures with improved polling logic and error detection mechanisms
  • Refined test prompts for better scenario specificity
  • Removed unused namespace configuration from test resources
  • Strengthened test pre-flight checks with comprehensive validation loops

Note: This release contains internal testing infrastructure updates with no user-facing feature changes.

_{✏️ Tip: You can customize this high-level summary in your review settings.}

[coderabbitai]

Caution

Review failed

The pull request is closed.

Walkthrough

Test fixtures for the Ask Holmes test suite are enhanced with more robust setup procedures. Changes include modifying test prompts, adding polling loops to verify pod readiness and log availability, detecting Kubernetes events, removing an unused namespace manifest, and adding cleanup procedures.

Changes

Cohort / File(s)	Summary
Test case 108 (logs_nearby_lines) tests/llm/fixtures/test_ask_holmes/108_logs_nearby_lines/manifest.yaml, test_case.yaml	Updated pod manifest to use infinite loop with 30-second sleeps instead of fixed 3600-second sleep. Enhanced before_test with polling loop (20 iterations, 3-second intervals) to verify pod readiness and detect two specific log lines; includes failure diagnostics dumping pod logs and listing pods.
Test case 16 (failed_no_toolset_found) tests/llm/fixtures/test_ask_holmes/16_failed_no_toolset_found/test_case.yaml	Refined user_prompt to specify target namespace (app-108) and clarify the question format.
Test case 17 (oom_kill) tests/llm/fixtures/test_ask_holmes/17_oom_kill/test_case.yaml	Extended user_prompt to include namespace specification. Added before_test polling loop (20 iterations, 3-second intervals, 60-second timeout) to detect Kubernetes OOMKilled events; includes detailed failure diagnostics with pod and event information.
Test case 23 (app_error_in_current_logs) tests/llm/fixtures/test_ask_holmes/23_app_error_in_current_logs/manifests/namespace.yaml, test_case.yaml	Deleted unused namespace manifest file. Enhanced before_test with ERROR_FOUND flag and polling loop (30 iterations, 3-second intervals) to verify "Failed to reach" error in curl pod logs; added failure path with diagnostics. Added namespace cleanup in after_test.

Estimated code review effort

🎯 2 (Simple) | ⏱️ ~10 minutes

Possibly related PRs

• add missing namespace cleanup to evals #752: Implements similar namespace deletion cleanup pattern in after_test blocks across multiple test fixtures.
• Fix more evals #718: Modifies the same test fixture (test_case_23) with overlapping changes to test setup and behavior verification.

Suggested reviewers

• aantn
• arikalon1
• moshemorad

---

📜 Recent review details

Configuration used: Organization UI

Review profile: CHILL

Plan: Pro

📥 Commits

Reviewing files that changed from the base of the PR and between 4548191 and 7b9aec8.

📒 Files selected for processing (6)

• tests/llm/fixtures/test_ask_holmes/108_logs_nearby_lines/manifest.yaml
• tests/llm/fixtures/test_ask_holmes/108_logs_nearby_lines/test_case.yaml
• tests/llm/fixtures/test_ask_holmes/16_failed_no_toolset_found/test_case.yaml
• tests/llm/fixtures/test_ask_holmes/17_oom_kill/test_case.yaml
• tests/llm/fixtures/test_ask_holmes/23_app_error_in_current_logs/manifests/namespace.yaml
• tests/llm/fixtures/test_ask_holmes/23_app_error_in_current_logs/test_case.yaml

---

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

[github-actions]

Results of HolmesGPT evals

• ask_holmes: 28/37 test cases were successful, 3 regressions, 2 setup failures, 3 mock failures

Test suite	Test case	Status
ask	01_how_many_pods	✅
ask	02_what_is_wrong_with_pod	✅
ask	04_related_k8s_events	✅
ask	05_image_version	✅
ask	09_crashpod	✅
ask	10_image_pull_backoff	✅
ask	110_k8s_events_image_pull	✅
ask	11_init_containers	✅
ask	13a_pending_node_selector_basic	✅
ask	14_pending_resources	✅
ask	15_failed_readiness_probe	✅
ask	163_compaction_follow_up	✅
ask	17_oom_kill	🚧
ask	18_oom_kill_from_issues_history	❌
ask	19_detect_missing_app_details	✅
ask	20_long_log_file_search	❌
ask	24_misconfigured_pvc	✅
ask	24a_misconfigured_pvc_basic	✅
ask	28_permissions_error	🚧
ask	39_failed_toolset	✅
ask	41_setup_argo	✅
ask	42_dns_issues_steps_new_tools	⚠️
ask	43_current_datetime_from_prompt	✅
ask	45_fetch_deployment_logs_simple	✅
ask	51_logs_summarize_errors	✅
ask	53_logs_find_term	✅
ask	54_not_truncated_when_getting_pods	✅
ask	59_label_based_counting	✅
ask	60_count_less_than	❌
ask	61_exact_match_counting	✅
ask	63_fetch_error_logs_no_errors	✅
ask	79_configmap_mount_issue	✅
ask	83_secret_not_found	✅
ask	86_configmap_like_but_secret	✅
ask	93_calling_datadog[0]	🔧
ask	93_calling_datadog[1]	🔧
ask	93_calling_datadog[2]	🔧

Legend

• ✅ the test was successful
• :minus: the test was skipped
• ⚠️ the test failed but is known to be flaky or known to fail
• 🚧 the test had a setup failure (not a code regression)
• 🔧 the test failed due to mock data issues (not a code regression)
• 🚫 the test was throttled by API rate limits/overload
• ❌ the test failed and should be fixed before merging the PR