Hardware Profiles

Hardware profiles are used to configure the hardware resources for the Cardano Rosetta Java application. They provide preconfigured sets of environment variables to optimize PostgreSQL and connection pooling based on different hardware capabilities.

Overview

Hardware profiles are implemented as environment files that are passed to Docker or Docker Compose when launching Cardano Rosetta Java services. The entrypoint script verifies these settings and configures PostgreSQL and application connection pools accordingly.

Available Hardware Profiles

Cardano Rosetta Java provides preconfigured hardware profiles to match different deployment scenarios:

Entry-Level

Entry-Level Hardware Profile

Designed for development environments or smaller deployments with limited resources:

• 4 CPU cores
• 32 GB RAM
• Suitable for development, testing, or light production use

Mid-Level

Mid-Level Hardware Profile

Optimized for production environments with moderate resources:

• 8 CPU cores
• 48 GB RAM
• Suitable for standard production deployments with moderate traffic

tip

Choose the mid-level profile for production environments with moderate to high traffic.

Hardware Profile Configuration Parameters

Hardware profiles define two main sets of configuration parameters:

Profile Parameters Comparison

Parameter	Entry-Level	Mid-Level	Purpose
API_DB_POOL_MIN_COUNT	12	150	Minimum database connections
API_DB_POOL_MAX_COUNT	12	150	Maximum database connections
DB_POSTGRES_MAX_CONNECTIONS	120	300	Maximum PostgreSQL connections
DB_POSTGRES_SHARED_BUFFERS	1GB	4GB	Memory for data caching
DB_POSTGRES_EFFECTIVE_CACHE_SIZE	2GB	8GB	Estimate of memory available for disk caching
DB_POSTGRES_WORK_MEM	16MB	64MB	Memory for query operations
DB_POSTGRES_MAINTENANCE_WORK_MEM	128MB	512MB	Memory for maintenance operations
DB_POSTGRES_WAL_BUFFERS	16MB	512MB	Memory for write-ahead logging
DB_POSTGRES_MAX_PARALLEL_WORKERS	4	8	Maximum parallel query workers

How to Use Hardware Profiles

Docker

docker run --name rosetta -v {MOUNT_PATH}:/node \
  --env-file ./docker/.env.dockerfile \
  --env-file ./docker/.env.docker-profile-mid-level \
  -p 8082:8082 --shm-size=4g -d cardanofoundation/cardano-rosetta-java:latest

Docker Compose

docker compose --env-file .env.docker-compose \
  --env-file .env.docker-compose-profile-mid-level \
  -f docker-compose.yaml up -d

Custom Hardware Profiles

You can create custom hardware profiles by:

1. Copying an existing profile file, e.g., docker/.env.docker-profile-mid-level
2. Modifying the parameters to match your hardware
3. Using the custom profile file when launching with Docker or Docker Compose

When creating custom profiles, consider:

• Maintain a healthy ratio between API_DB_POOL_*_COUNT and DB_POSTGRES_MAX_CONNECTIONS.
• Adjust memory parameters (DB_POSTGRES_SHARED_BUFFERS, DB_POSTGRES_EFFECTIVE_CACHE_SIZE, DB_POSTGRES_WORK_MEM, etc.) based on your system's available RAM.
• Set DB_POSTGRES_MAX_PARALLEL_WORKERS according to your CPU core count.

info

Hardware profiles help optimize PostgreSQL and connection pooling based on your available resources.