AI in Godalming
- AI in Godalming: Server Configuration Documentation
This document details the server configuration for the "AI in Godalming" project, a localized artificial intelligence initiative focused on data analysis for local council services. This guide is intended for new system administrators and developers contributing to the project. Please read this document thoroughly before making any changes to the server environment. Familiarity with Linux server administration and MediaWiki administration is highly recommended.
Overview
The "AI in Godalming" project utilizes a cluster of servers hosted in a dedicated data center in Guildford. The primary goal is to process and analyze publicly available data, combined with anonymized council data, to improve resource allocation and service delivery. The system leverages machine learning algorithms, requiring considerable processing power and storage capacity. This document will cover the hardware, software, and network configuration of these servers. We utilize Debian Linux as our base operating system for stability and security.
Hardware Specifications
The server cluster consists of three main types of servers: Database Servers, Application Servers, and Processing Nodes. The following tables detail the specifications of each type.
| Server Type | CPU | RAM | Storage | Network Interface |
|---|---|---|---|---|
| Database Servers (2x) | Intel Xeon Gold 6248R (24 cores) | 256 GB DDR4 ECC | 4 x 4TB NVMe SSD (RAID 10) | 10 Gbps Ethernet |
| Application Servers (3x) | AMD EPYC 7543P (32 cores) | 128 GB DDR4 ECC | 2 x 2TB NVMe SSD (RAID 1) + 8TB HDD | 1 Gbps Ethernet |
| Processing Nodes (5x) | NVIDIA A100 GPU + Intel Xeon Silver 4210 (10 cores) | 64 GB DDR4 ECC | 1 x 1TB NVMe SSD | 10 Gbps Ethernet + 10 Gbps InfiniBand |
Regular hardware monitoring is performed using Nagios to ensure optimal performance and identify potential failures. Access to the hardware is restricted to authorized personnel only; see the Security Policy for details.
Software Configuration
All servers are running Debian 11 (Bullseye) with a customized kernel optimized for machine learning workloads. The following software components are crucial to the system's operation.
| Software Component | Version | Server Type | Purpose |
|---|---|---|---|
| PostgreSQL | 14.5 | Database Servers | Data storage and retrieval. |
| Python 3 | 3.9.7 | All Servers | Primary scripting language for AI models. |
| TensorFlow | 2.8.0 | Processing Nodes | Machine learning framework. |
| PyTorch | 1.12.1 | Processing Nodes | Alternative machine learning framework. |
| Nginx | 1.21.3 | Application Servers | Web server and reverse proxy. |
| Gunicorn | 20.1.0 | Application Servers | Python WSGI HTTP server. |
Software updates are managed using APT package management. A robust backup strategy is in place, utilizing daily full backups and hourly incremental backups to a remote storage location. Configuration management is handled with Ansible to ensure consistency across all servers.
Network Configuration
The server cluster is connected to the internal network via a dedicated VLAN. A firewall, configured using iptables, protects the servers from unauthorized access.
| Network Component | IP Address Range | Purpose |
|---|---|---|
| VLAN 100 (AI Cluster) | 192.168.100.0/24 | Internal network for AI servers. |
| Database Servers | 192.168.100.10-11 | Host PostgreSQL databases. |
| Application Servers | 192.168.100.20-22 | Serve web applications and APIs. |
| Processing Nodes | 192.168.100.30-34 | Perform machine learning computations. |
| Load Balancer | 192.168.100.1 | Distributes traffic to application servers. |
Access to the servers from the external network is restricted to specific ports and IP addresses, as defined in the Firewall Rules. The load balancer distributes traffic across the application servers to ensure high availability and scalability. We use DNS records to map domain names to the load balancer's IP address. The network is monitored using Zabbix for performance and security.
Security Considerations
Security is paramount. All servers are hardened according to the CIS Benchmarks. Regular security audits are conducted to identify and address vulnerabilities. User access is controlled via SSH keys and multi-factor authentication. All data is encrypted both in transit and at rest. The Incident Response Plan outlines procedures for handling security breaches.
Main Page Server Administration Database Administration Network Security Machine Learning Data Analysis System Monitoring Backup and Recovery Configuration Management Firewall Configuration Security Policy Incident Response Plan APT package management Nagios Zabbix DNS records IPTables
Intel-Based Server Configurations
| Configuration | Specifications | Benchmark |
|---|---|---|
| Core i7-6700K/7700 Server | 64 GB DDR4, NVMe SSD 2 x 512 GB | CPU Benchmark: 8046 |
| Core i7-8700 Server | 64 GB DDR4, NVMe SSD 2x1 TB | CPU Benchmark: 13124 |
| Core i9-9900K Server | 128 GB DDR4, NVMe SSD 2 x 1 TB | CPU Benchmark: 49969 |
| Core i9-13900 Server (64GB) | 64 GB RAM, 2x2 TB NVMe SSD | |
| Core i9-13900 Server (128GB) | 128 GB RAM, 2x2 TB NVMe SSD | |
| Core i5-13500 Server (64GB) | 64 GB RAM, 2x500 GB NVMe SSD | |
| Core i5-13500 Server (128GB) | 128 GB RAM, 2x500 GB NVMe SSD | |
| Core i5-13500 Workstation | 64 GB DDR5 RAM, 2 NVMe SSD, NVIDIA RTX 4000 |
AMD-Based Server Configurations
| Configuration | Specifications | Benchmark |
|---|---|---|
| Ryzen 5 3600 Server | 64 GB RAM, 2x480 GB NVMe | CPU Benchmark: 17849 |
| Ryzen 7 7700 Server | 64 GB DDR5 RAM, 2x1 TB NVMe | CPU Benchmark: 35224 |
| Ryzen 9 5950X Server | 128 GB RAM, 2x4 TB NVMe | CPU Benchmark: 46045 |
| Ryzen 9 7950X Server | 128 GB DDR5 ECC, 2x2 TB NVMe | CPU Benchmark: 63561 |
| EPYC 7502P Server (128GB/1TB) | 128 GB RAM, 1 TB NVMe | CPU Benchmark: 48021 |
| EPYC 7502P Server (128GB/2TB) | 128 GB RAM, 2 TB NVMe | CPU Benchmark: 48021 |
| EPYC 7502P Server (128GB/4TB) | 128 GB RAM, 2x2 TB NVMe | CPU Benchmark: 48021 |
| EPYC 7502P Server (256GB/1TB) | 256 GB RAM, 1 TB NVMe | CPU Benchmark: 48021 |
| EPYC 7502P Server (256GB/4TB) | 256 GB RAM, 2x2 TB NVMe | CPU Benchmark: 48021 |
| EPYC 9454P Server | 256 GB RAM, 2x2 TB NVMe |
Order Your Dedicated Server
Configure and order your ideal server configuration
Need Assistance?
- Telegram: @powervps Servers at a discounted price
⚠️ *Note: All benchmark scores are approximate and may vary based on configuration. Server availability subject to stock.* ⚠️