SIENTIAPDE-1350: Implement file cleanup workflow and activities, including MinIO and local directory cleanup, configuration, and metrics.
This commit is contained in:
83
.env.example
83
.env.example
@@ -1,44 +1,57 @@
|
||||
POSTGRES_HOST="paradedb-rw.paradedb.svc.cluster.local"
|
||||
POSTGRES_PORT="5432"
|
||||
POSTGRES_USER="sientia"
|
||||
POSTGRES_PASSWORD="password"
|
||||
POSTGRES_DBNAME="sientia"
|
||||
POSTGRES_MIN_CONNECTIONS="10"
|
||||
POSTGRES_MAX_CONNECTIONS="30"
|
||||
POSTGRES_HOST=paradedb-rw.paradedb.svc.cluster.local
|
||||
POSTGRES_PORT=5432
|
||||
POSTGRES_USER=sientia
|
||||
POSTGRES_PASSWORD=password
|
||||
POSTGRES_DBNAME=sientia
|
||||
POSTGRES_MIN_CONNECTIONS=10
|
||||
POSTGRES_MAX_CONNECTIONS=30
|
||||
|
||||
MLFLOW_URL="http://sientia-tracker-mlflow-tracking.sientia-tracker.svc.cluster.local:80"
|
||||
MLFLOW_USERNAME="aignosi"
|
||||
MLFLOW_PASSWORD="mlflow_password"
|
||||
MLFLOW_URL=http://sientia-tracker-mlflow-tracking.sientia-tracker.svc.cluster.local:80
|
||||
MLFLOW_USERNAME=aignosi
|
||||
MLFLOW_PASSWORD=mlflow_password
|
||||
|
||||
LOG_LEVEL="DEBUG"
|
||||
HTTP_METRICS_PORT="9090"
|
||||
HTTP_SDK_METRICS_PORT="9091"
|
||||
PROJECT_NAME="sientia-model-manager"
|
||||
LOG_LEVEL=DEBUG
|
||||
HTTP_METRICS_PORT=9090
|
||||
HTTP_SDK_METRICS_PORT=9091
|
||||
PROJECT_NAME=sientia-model-manager
|
||||
|
||||
TEMPORAL_HOST="temporal-frontend.temporal.svc.cluster.local:7233"
|
||||
TEMPORAL_NAMESPACE="model-manager"
|
||||
TEMPORAL_HOST=temporal-frontend.temporal.svc.cluster.local:7233
|
||||
TEMPORAL_NAMESPACE=model-manager
|
||||
|
||||
MONGODB_USERNAME="mongo_user"
|
||||
MONGODB_PASSWORD="mongo_db_password"
|
||||
MONGODB_URL="my-release-mongodb.mongodb.svc.cluster.local:27017"
|
||||
MONGODB_DATABASE="sientia"
|
||||
MONGODB_TTL_INDEX_HOURS="1"
|
||||
MONGODB_USERNAME=mongo_user
|
||||
MONGODB_PASSWORD=mongo_db_password
|
||||
MONGODB_URL=my-release-mongodb.mongodb.svc.cluster.local:27017
|
||||
MONGODB_DATABASE=sientia
|
||||
MONGODB_TTL_INDEX_HOURS=1
|
||||
|
||||
MINIO_ENDPOINT_URL="http://minio.minio.svc.cluster.local:9000"
|
||||
MINIO_ACCESS_KEY="minioadmin"
|
||||
MINIO_SECRET_KEY="minioadmin"
|
||||
MINIO_REGION="us-east-1"
|
||||
MINIO_USE_SSL="false"
|
||||
MINIO_MAX_RETRY_ATTEMPTS="3"
|
||||
MINIO_RETRY_MODE="adaptive"
|
||||
MINIO_CONNECT_TIMEOUT="10"
|
||||
MINIO_READ_TIMEOUT="60"
|
||||
MINIO_ENDPOINT_URL=http://minio.minio.svc.cluster.local:9000
|
||||
MINIO_ACCESS_KEY=minioadmin
|
||||
MINIO_SECRET_KEY=minioadmin
|
||||
MINIO_REGION=us-east-1
|
||||
MINIO_USE_SSL=false
|
||||
MINIO_MAX_RETRY_ATTEMPTS=3
|
||||
MINIO_RETRY_MODE=adaptive
|
||||
MINIO_CONNECT_TIMEOUT=10
|
||||
MINIO_READ_TIMEOUT=60
|
||||
|
||||
# Workflow Activity Timeouts (in seconds)
|
||||
# These timeouts are designed to handle large files (up to 200MB)
|
||||
TIMEOUT_VALIDATE_PARAMS="30" # Parameter validation (fast operation)
|
||||
TIMEOUT_TRAIN_MODEL="2700" # Model training (30 min for large datasets)
|
||||
TIMEOUT_DELETE_FILE="120" # Delete file from MinIO (1 min)
|
||||
TIMEOUT_UPDATE_DATABASE="30" # Database update operations (30 sec)
|
||||
TIMEOUT_VALIDATE_PARAMS=30 # Parameter validation (fast operation)
|
||||
TIMEOUT_TRAIN_MODEL=2700 # Model training (30 min for large datasets)
|
||||
TIMEOUT_DELETE_FILE=120 # Delete file from MinIO (1 min)
|
||||
TIMEOUT_UPDATE_DATABASE=30 # Database update operations (30 sec)
|
||||
|
||||
EXTRA_PIP_REQUIREMENTS="git+https://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@github.com/Aignosi/sientia-mlops-library.git"
|
||||
# Cleanup Configuration
|
||||
# Cleanup retention period in hours (files older than this will be deleted)
|
||||
CLEANUP_RETENTION_HOURS=24 # Default: 24 hours
|
||||
# Enable dry run mode to test without actually deleting files
|
||||
CLEANUP_DRY_RUN=false # Set to true for testing without deletion
|
||||
# Cleanup operation timeouts
|
||||
TIMEOUT_CLEANUP_MINIO=300 # MinIO cleanup timeout (5 minutes)
|
||||
TIMEOUT_CLEANUP_LOCAL=120 # Local directory cleanup timeout (2 minutes)
|
||||
# MinIO list operation page size for cleanup
|
||||
MAX_KEYS_CLEANUP=1000 # Maximum keys per page when listing objects
|
||||
# Default bucket for cleanup operations
|
||||
DEFAULT_CLEANUP_BUCKET=model-training # Default bucket to clean
|
||||
|
||||
EXTRA_PIP_REQUIREMENTS=git+https://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@github.com/Aignosi/sientia-mlops-library.git
|
||||
|
||||
Reference in New Issue
Block a user