Server rental store

AI in the Samoan Rainforest

AI in the Samoan Rainforest: Server Configuration

This article details the server configuration for the “AI in the Samoan Rainforest” project, designed to process data from a network of sensors deployed throughout the rainforests of Samoa. This project utilizes artificial intelligence to analyze biodiversity, track invasive species, and monitor environmental changes. This document is aimed at new system administrators joining the team.

Project Overview

The “AI in the Samoan Rainforest” project relies on real-time data analysis of audio, visual, and environmental sensor data. Data is collected from remote sensor nodes and transmitted to a central server farm for processing. The server infrastructure is crucial for the project's success, requiring high availability, scalability, and robust data storage. Understanding the individual components and their interdependencies is key to effective maintenance and troubleshooting. We utilize a hybrid cloud approach, leveraging both on-premise hardware and cloud services for optimal performance and cost-effectiveness. See also Data Acquisition Protocols and Sensor Network Topology.

Server Hardware Specification

The core of the server infrastructure consists of three primary server types: Data Ingestion Servers, Processing Servers, and Database Servers. The following tables detail the specifications for each:

Server Type CPU RAM Storage Network Interface
Data Ingestion Servers | Intel Xeon Silver 4310 (12 Cores) | 64GB DDR4 ECC | 4TB NVMe SSD (RAID 1) | 10Gbps Ethernet |
Processing Servers | AMD EPYC 7763 (64 Cores) | 256GB DDR4 ECC | 8TB NVMe SSD (RAID 0) + 16TB HDD (RAID 5) | 25Gbps Ethernet |
Database Servers | Intel Xeon Gold 6338 (32 Cores) | 128GB DDR4 ECC | 16 x 4TB SAS HDD (RAID 6) | 10Gbps Ethernet |

These servers are housed in a dedicated, climate-controlled server room located at the research facility in Apia. Power redundancy is provided by a UPS system with a minimum of 30 minutes of backup power. A detailed inventory of all hardware is maintained in the Hardware Inventory Database.

Software Stack

The software stack is designed for efficient data processing and storage. The following table summarizes the key software components:

Component Version Purpose Operating System
Operating System | Ubuntu Server 22.04 LTS | Base operating system for all servers | Ubuntu |
Data Ingestion | Nginx | Reverse proxy and load balancer | Ubuntu |
Data Processing | Python 3.9 with TensorFlow 2.8 | AI model execution and data analysis | Ubuntu |
Database | PostgreSQL 14 | Data storage and retrieval | Ubuntu |
Monitoring | Prometheus & Grafana | System monitoring and visualization | Ubuntu |
Containerization | Docker & Kubernetes | Application deployment and orchestration | Ubuntu |

All code is version controlled using Git Repository Access. Deployment is automated using Continuous Integration/Continuous Deployment (CI/CD) Pipeline. Regular security audits are performed as per the Security Policy.

Network Configuration

The server network is segmented into three VLANs: Data Ingestion, Processing, and Database. This segmentation enhances security and improves network performance. Firewall rules are configured to restrict communication between VLANs to only necessary services.

VLAN ID Subnet Purpose Gateway
10 | 192.168.10.0/24 | Data Ingestion Servers | 192.168.10.1 |
20 | 192.168.20.0/24 | Processing Servers | 192.168.20.1 |
30 | 192.168.30.0/24 | Database Servers | 192.168.30.1 |

External access is provided through a dedicated internet connection with a static IP address. DNS records are managed internally using Internal DNS Configuration. Network diagrams are available in the Network Topology Documentation.

Data Flow

1. Sensor data is transmitted to the Data Ingestion Servers. 2. Nginx load balances the incoming data across multiple Data Ingestion Servers. 3. Data Ingestion Servers validate and pre-process the data. 4. Pre-processed data is sent to the Processing Servers via the network. 5. Processing Servers execute AI models to analyze the data. 6. Analyzed data is stored in the Database Servers. 7. Researchers access the data through a web interface. See Data Access Procedures.

Future Considerations

Future enhancements to the server infrastructure may include:

⚠️ *Note: All benchmark scores are approximate and may vary based on configuration. Server availability subject to stock.* ⚠️

View full article | Back to Server rental store