QTZPOC-20: Enable code_paths for sklearn models; improve worker shutdown handling
The MLFlowRepository now resolves and forwards code_paths to mlflow.sklearn.log_model, ensuring custom code dependencies for sklearn models are preserved and correctly reloaded. A new integration test validates this end-to-end functionality. Additionally, worker exception handling is refined to distinguish between deliberate shutdowns (SystemExit, KeyboardInterrupt) and unhandled exceptions, preventing unnecessary error logging for expected exits.
QTZPOC-20: Refactor ModelImport status engine for lazy initialization, add byte-counted size validation to bundle extraction, optimize MLflow run ID lookups, and ensure workflow cancellation propagates correctly.
QTZPOC-20: Refactor model import workflow, introduce atomic model ID generation, and clarify MinIO expiry.
Fuses bundle download, decryption, MLflow experiment creation, and artifact upload into a single activity to prevent cross-worker issues. Implements an atomic counter for MongoDB model IDs, replacing the max(id) + 1 scan to avoid collisions during concurrent imports. Adds comprehensive documentation for MinIO object expiry mechanisms and refines lifecycle rule application. Improves error handling with clearer responsibility classification and centralizes MLflow repository injection.
QTZPOC-20: Enhance system robustness and test coverage across bundle processing, activity error handling, and worker operations.
* Strengthen bundle reader validation with new tests for archive integrity, security vulnerabilities (e.g., path traversal, duplicate entries, special file types), and I/O error resilience.
* Improve error handling and status propagation in PI Web API, MLflow, and OPC activities, including partial write scenarios and OPC session management.
* Add comprehensive tests for worker startup, configuration, graceful shutdown, and Prometheus metrics server initialization.
* Introduce a new utility for intelligent DataFrame debugging, summarizing large dataframes for concise logging, and refine exception chain summarization.
* Update code coverage configuration to exclude @overload decorator lines.
QTZPOC-20: Refactor model repository tests for improved assertion logic and exception handling, including correcting pytest.raises exception access and refining mock verifications.
QTZPOC-20: Prioritize code/ directory download and sys.path injection before model loading, enabling MLflow flavors to stream models directly from URIs and simplifying joblib artifact handling.
QTZPOC-20: Unify artifact code directory handling for all model flavors and standardize joblib pickle path
Implemented a robust mechanism to ensure an artifact's code/ directory is prepended to sys.path before loading for sklearn, pyfunc, and joblib flavors. sklearn and pyfunc now resolve MLflow URIs to local directories first to facilitate this sys.path modification. Standardized the joblib flavor to explicitly read and write model.pkl within the artifact directory. Updated documentation and added comprehensive tests to reflect these changes.
QTZPOC-20: Improve structured logging and reporting for all model import pipeline failures by introducing a dedicated activity and centralizing error classification. This ensures consistent observability, with all workflow failures now reported before cleanup. Also, adjusted import step timeouts and refined manual run script cleanup logic.
QTZPOC-20: Implement explicit joblib model flavor and enhance model import workflow with extensive E2E tests.
This commit introduces a first-class joblib model flavor for direct loading and saving of Python pickle artifacts, including custom code within the artifact's code/ directory, bypassing MLflow's flavor loaders. The model import workflow's status database configuration is now streamlined to default to the POSTGRES_* server, with explicit IMPORT_STATUS_DB_* overrides.
Comprehensive end-to-end tests have been added for the model import workflow, covering success paths, gate rejections, digest enforcement, provisioning failures, status record integrity, cleanup, and retry/recovery scenarios using a real-clock environment. New E2
QTZPOC-20: Implement model import workflow for .sientia bundles
Introduces the import_model workflow and its associated activities to enable importing .sientia bundles into MLflow and MongoDB. This includes:
* New core workflow and activity logic for handling bundle download, decryption, validation, MLflow provisioning (experiments, runs, registered versions), and MongoDB model document creation.
* Comprehensive E2E and unit test suites, including a dedicated test harness and a manual run script for development.
* Updated .env.example and README.md with new configurations and usage instructions.
* Ensures correct MLflow model source resolution for legacy loader compatibility and robust MongoDB document management.
* Refactors existing utility scripts into a scripts/ directory.
QTZPOC-20: Implement .sientia model bundle import workflow
Introduces a new Temporal workflow and activities to import .sientia model bundles. This enables:
- Secure decryption and validation of bundles through a seven-gate process.
- Provisioning imported models in MLflow and updating the MongoDB model listing.
- Comprehensive status tracking in the public.experiment_run table with new IMPORT_SUCCESS/IMPORT_ERROR states.
- Dedicated metrics, robust error handling, and configurable settings.
- Includes updated e2e database fixtures and extensive unit/integration tests for bundle format, reader, and overall workflow.
QTZPOC-13: Centralize MLflow joblib fallback logic, ensuring consistent model version resolution by reusing the initial run ID, improving logging, and updating dependencies.
QTZPOC-13: Introduce joblib fallback for MLflow model loading
Enhances model loading resilience by attempting to deserialize models directly with joblib if MLflow's native flavor loaders fail. This mechanism handles cases where .pkl files are present in the artifact.
Adds joblib as a core dependency and includes new end-to-end and unit tests to validate fallback behavior, covering success, no-fallback, and double-failure scenarios. Also updates .gitignore for .claude/ files and incorporates minor refactorings and linting adjustments across E2E tests and utility functions.
Renamed the OPC.close asynchronous method to OPC.aclose to align with common Python conventions for asynchronous context managers and methods, improving clarity. All call sites and tests have been updated accordingly.
Enhance MLFlow run ID resolution with error handling for missing and invalid source URIs
- Added checks in `get_model_run_id` method to raise exceptions for models with missing or invalid source URIs.
- Introduced new test cases to validate error handling for these scenarios.
- Updated `requirements-light.txt` to include `mlflow` as a dependency.
Update .gitignore, requirements, and enhance OPC UA error handling
- Added new entries to .gitignore for openspec and cursor directories.
- Updated sientia-dataops-library dependency version in requirements-light.txt from 1.10.4 to 1.12.0.
- Enhanced OPC UA communication by refining reconnect logic and error handling in opc_repository.py, including the introduction of a reconnect flag and improved session management.
- Updated tests to cover new reconnect scenarios and ensure robust error handling for protocol states.
SIENTIAPDE-1811 Implement OPC write error handling and refactor tag writing logic
- Introduced a new function `_apply_opc_write_error` to manage session and reconnect flags based on OPC write error responses.
- Refactored the `_write_tags_from_config` method to streamline the writing of OPC tags for both prediction and confidence data.
- Enhanced unit tests to cover various scenarios for OPC write errors, including session bad and reconnect in progress cases.
- Updated existing tests to validate the new logic and ensure robust error handling.
Enhance OPC UA communication and metrics tracking
- Updated README.md to include new OPC UA Communication section and detailed metrics for session and write diagnostics.
- Added new metrics in laborious/metrics.py for tracking OPC UA session states and write attempts.
- Refactored OPC activity in laborious/activities/opc.py to handle session errors and improve error reporting.
- Updated e2e tests to cover new scenarios for OPC session/channel errors and reconnect handling.
- Modified .gitignore to include relatorio files and mlruns directory.
- Added ipykernel to requirements-dev.txt for Jupyter notebook support.
SIENTIAPDE-1712 Enhance logging across various classes by adding logger parameters and improving debug statements. This update includes adjustments in Gates, MLFlow, ModelMetrics, and MLFlowRepository classes for better traceability and observability during operations.
Update dependencies and refactor input filter handling for consistency
- Updated sientia-dataops-library dependency version from 1.10.3 to 1.10.4 in requirements.txt.
- Refactored input filter handling in the Gates class to read policy and config keys in a case-insensitive manner.
- Updated test cases to ensure consistency in filter key naming conventions across various scenarios.
Refactor metrics and API handling for improved consistency and clarity
- Removed the SIENTIA_CORE_LABELS constant and replaced it with CORE_LABELS for uniformity across metrics.
- Updated the API class to ensure operation_type is always included in core labels for PI Web API metrics.
- Simplified metric tag handling in the Gates class by consolidating common tags into a single core_tags dictionary.
- Enhanced the PredictionProcess class to improve error handling and variable naming for clarity.
Update environment variables in values.yaml and enhance metric labels in metrics.py
- Changed POSTGRES_USER and POSTGRES_PASSWORD values in values.yaml for improved security.
- Added 'runtime' label to metrics in metrics.py for better environment identification.
- Updated CORE_LABELS to include 'runtime' for consistency across metrics.
- Modified type hint for data parameter in PredictionProcess to use a dictionary for better clarity.
- Adjusted tests to reflect changes in core labels and MinIO configuration.
Refactor MinioDataFramePayload usage across activities
- Updated instances of MinioDataFramePayload initialization in Gates, MLFlow, and Storage classes to use the new from_dict method for better data reconstruction from dictionaries.
- Enhanced the PredictionProcess workflow to utilize the updated payload handling.
- Added passthrough fixtures in tests to accommodate the new from_dict method for consistent testing behavior.
Enhance README and Implement Drift Detection and Metrics Workflows
- Added new sections in README for Drift Workflow and Simple Metrics Workflow, detailing their execution flows and functionalities.
- Introduced `drift.py` for data drift detection, comparing current data against reference datasets.
- Added `simple_metrics.py` for calculating regression metrics (RMSE, MSE, MAE, R²).
- Updated `values.yaml` to include configuration for MinIO retention hours and offload threshold.
- Refactored `minio_dataframe_payload.py` to use the new offload threshold environment variable.
- Adjusted tests to reflect changes in environment variable handling for MinIO offload threshold.
Remove code validation script and refactor imports in activities and workflows
- Deleted the `validate.sh` script, which was responsible for running code quality checks.
- Cleaned up import statements in `activities.py`, `gates.py`, `mlflow.py`, and `storage.py` by removing unused imports and organizing them.
- Refactored initialization methods in `MinioManager` and `MLFlow` classes for improved readability.
- Updated various workflows to ensure compatibility with the new structure and removed unnecessary comments.
- Enhanced test cases to accommodate changes in the activities and workflows, ensuring proper mocking of dependencies.
Implement MinIO Offload and Retention Features
- Added configuration options for MinIO retention hours and offload threshold in README.
- Introduced MinIO payload offloading for large DataFrame-derived payloads, storing them as parquet files.
- Updated activities to utilize MinIO for data loading and cleanup, including new methods for offloading and retention management.
- Refactored existing activities to integrate MinIO functionality, ensuring compatibility with previous workflows.
- Removed the legacy MinioRepository class, consolidating MinIO operations under a new manager structure.
- Updated requirements to use the latest version of the sientia-dataops-library.
Refactor PI Web API response handling in tests
- Updated test cases in test_api.py to handle response data as lists instead of dictionaries for consistency with the API's expected output format.
- Adjusted mock responses to reflect the new structure, ensuring tests accurately simulate API behavior.
- Enhanced clarity in test descriptions and improved overall test coverage for response processing scenarios.
Update sonar-project.properties to exclude all worker files from coverage and modify execution counts and timestamps in tests.ipynb. Add a new test for empty DataFrame handling in test_model_repository.py.
Refactor validation script and improve logging in API and model repository
- Updated validation script to include 'e2e/' directory in code formatting and linting checks.
- Enhanced error logging in API class to improve readability of error messages.
- Refactored debug logging in model repository for better structured output.
- Cleaned up import statements in various files for improved organization.
Enhance end-to-end tests for PredictionsBatch workflow scenarios
- Introduced mock repositories for PI Web API and OPC operations to improve test coverage.
- Updated test scenarios to handle partial write errors for PI Web API and OPC.
- Refactored existing tests to assert correct behavior under various error conditions.
- Enhanced logging and error handling in API and OPC activities to provide clearer feedback on failures.
- Removed outdated integration test file to streamline test suite.
SIENTIAPDE-1478
Enhance end-to-end tests for PredictionsBatch workflow
- Added new test scenarios for input and transform gates handling CONTINUE, STOP, and REPEAT policies.
- Implemented sample data insertion functions for testing various prediction outcomes.
- Updated existing tests to verify behavior under different input conditions and response validations.
- Refactored test structure for clarity and maintainability.
Update coverage source in pyproject.toml, add testcontainers for PostgreSQL in requirements-dev.txt, increment image tag and adjust probe delays in values.yaml, and refine condition checks in format_and_export_prediction.py and mlflow.py. Additionally, enhance test coverage in test_gates.py.
