venice_ai.presets.production

Production configuration preset for Venice AI SDK.

This preset provides battle-tested configuration optimized for production deployments with enterprise-grade features including:

• Redis backend for distributed state management
• Intelligent scheduler with rate limiting
• Circuit breaker protection
• Optimized connection pooling
• Conservative timeouts for reliability

create_production_config

def create_production_config(
        redis_url: str | None = None,
        redis_key_prefix: str = "venice:prod:",
        max_concurrent_executions: int = 100,
        max_queue_size: int = 5000,
        enable_metrics: bool = True,
        _allow_localhost_for_testing: bool = False) -> VeniceAIConfig

Create a production-optimized configuration.

This configuration is designed for production deployments and includes:

• Redis backend for distributed state across multiple instances
• Intelligent scheduler with automatic rate limit discovery
• Circuit breaker with conservative thresholds
• Connection pooling optimized for high throughput
• Retry logic with exponential backoff

Arguments:

• redis_url - Redis connection URL. REQUIRED for production use. Set via VENICE_REDIS_URL environment variable or pass explicitly.
• Example - redis://user:pass@prod-redis:6379/0 CRITICAL: Using localhost in production will cause failures in distributed/multi-instance environments.
• redis_key_prefix - Prefix for all Redis keys to avoid collisions (default: venice:prod:)
• max_concurrent_executions - Maximum concurrent requests (default: 100)
• max_queue_size - Maximum queue size before rejecting requests (default: 5000)
• enable_metrics - Enable metrics collection (default: True)

Returns:

VeniceAIConfig configured for production use

Raises:

• ValueError - If redis_url is not provided via parameter or VENICE_REDIS_URL environment variable

Example:

>>> import os
>>> os.environ['VENICE_REDIS_URL'] = 'redis://prod-redis:6379'
>>> from venice_ai.presets import create_production_config
>>> from venice_ai import VeniceClient
>>>
>>> config = create_production_config(
...     max_concurrent_executions=200
... )
>>> client = VeniceClient(config=config, api_key="your-key")

Best Practices:

• ALWAYS set VENICE_REDIS_URL environment variable in production
• Use a dedicated Redis instance for production
• Monitor Redis connection pool metrics
• Adjust max_concurrent_executions based on your rate limits
• Enable metrics for production observability
• Use separate Redis key prefixes for different environments

create_production_config_high_throughput

def create_production_config_high_throughput(
        redis_url: str | None = None,
        redis_key_prefix: str = "venice:prod:",
        _allow_localhost_for_testing: bool = False) -> VeniceAIConfig

Create a high-throughput production configuration.

Optimized for maximum throughput when you have high rate limits. Uses more aggressive settings than standard production config.

Arguments:

• redis_url - Redis connection URL. REQUIRED for production use. Set via VENICE_REDIS_URL environment variable or pass explicitly. CRITICAL: Using localhost in production will cause failures in distributed/multi-instance environments.
• redis_key_prefix - Prefix for Redis keys

Returns:

VeniceAIConfig optimized for high throughput

Raises:

• ValueError - If redis_url is not provided via parameter or VENICE_REDIS_URL environment variable

Warnings:

Only use this preset if you have sufficient rate limits and infrastructure to handle high concurrency.

create_production_config_conservative

def create_production_config_conservative(
        redis_url: str | None = None,
        redis_key_prefix: str = "venice:prod:",
        _allow_localhost_for_testing: bool = False) -> VeniceAIConfig

Create a conservative production configuration.

Optimized for maximum reliability over throughput. Use this when stability is more important than speed.

Arguments:

• redis_url - Redis connection URL. REQUIRED for production use. Set via VENICE_REDIS_URL environment variable or pass explicitly. CRITICAL: Using localhost in production will cause failures in distributed/multi-instance environments.
• redis_key_prefix - Prefix for Redis keys

Returns:

VeniceAIConfig optimized for reliability

Raises:

• ValueError - If redis_url is not provided via parameter or VENICE_REDIS_URL environment variable