Mikatsu Technical Infrastructure & Systems Architecture Report
Report Date: August 17, 2025
Classification: Engineering Overview
1.0 Executive Summary
This document provides a detailed technical overview of the Mikatsu hosting infrastructure, a multi-tenant architecture designed for high-performance, security, and operational efficiency. The architecture is built upon three core pillars: a strategic partnership with a premier Infrastructure-as-a-Service (IaaS) provider, Interxion: A Digital Realty Company, for all physical data center facilities; a commitment to sustainable and efficient operations (Green Hosting); and a standardized service delivery framework leveraging SitePad for rapid, scalable website deployment.
The resulting platform is an optimized solution for delivering managed web services at scale. It combines the resilience and global reach of a Tier III+ data center network with a finely tuned software stack, designed to minimize server-side processing overhead and maximize resource utilization. This report will dissect each layer of the infrastructure, from the physical facilities to the application delivery stack.
2.0 Physical Infrastructure: Interxion/Digital Realty Data Centers
Mikatsu's entire physical footprint is co-located within Interxion/Digital Realty facilities. This strategic decision abstracts the physical layer management (power, cooling, security, connectivity) to a specialized global leader, allowing Mikatsu to focus on its service and software layers.
Subsystem Specification | Technical Implementation & Engineering Benefits |
Power Redundancy | 2N Fully Redundant Power Architecture. Each server rack is fed by two independent A/B power distribution units (PDUs), which are connected to separate UPS systems, switchgear, and utility feeds. The UPS systems are double-conversion online models ensuring zero transfer time. In case of a utility failure, N+1 configured diesel generators with priority refueling contracts can sustain the full data center load indefinitely. This design meets or exceeds Tier III+ uptime standards. |
Physical Security | Multi-Layered, Zero-Trust Security Model. |
Network Connectivity | Carrier-Neutral, BGP-Routed Network. |
Global Scalability | Standardized Global Platform. |
Compliance & Audits | Third-Party Audited Certifications. |
3.0 Global Network Topology & Data Center Locations
Region | Primary Metro Hub | Strategic Engineering Advantage |
North America (4) | Ashburn, VA, USA | Located in "Data Center Alley," with direct peering to the world's largest concentration of fiber backbones (MAE-East). Central US location providing balanced, low-latency routes to both East and West coasts. Proximity to major tech HQs and Pacific-facing subsea cable landing stations. Key Canadian business hub with robust national and cross-border connectivity. |
Europe (6) | Frankfurt, Germany Amsterdam, Netherlands London, UK Paris, France Madrid, Spain Stockholm, Sweden | Home to DE-CIX, the world's largest IXP by traffic volume, ensuring optimal peering across Europe. Another critical IXP hub (AMS-IX) with extensive global network reach. Premier financial hub with the highest density of transatlantic subsea cable connectivity. Strategic access point for Western European markets. Key gateway connecting Europe to North Africa and South America via subsea cables. Highly efficient and sustainable data center market serving the Nordic region. |
Asia-Pacific (4) | Singapore Tokyo, Japan Sydney, Australia Hong Kong | The primary, low-latency network gateway for the entire Southeast Asian market. A major financial and technology hub with one of the most advanced network infrastructures in Asia. The main interconnection point for the Australian market. A strategic, low-latency gateway to mainland China and other Asian markets. |
South America (1) | São Paulo, Brazil | The largest peering ecosystem in South America, serving the region's biggest economy. |
4.0 Service Delivery: Software & Hardware Stack
Component |
| Technical Purpose & Performance Impact | |
Operating System |
| Implements kernel-level virtualization to create Lightweight Virtualized Environments (LVEs) for each tenant. The LVE Manager strictly enforces resource allocations (CPU, IOPS, memory, processes), preventing "noisy neighbor" scenarios. CageFS virtualizes the filesystem, completely isolating users and preventing information disclosure vulnerabilities. | |
Web Server | LiteSpeed Web Server (LSWS) | An event-driven, high-performance web server that significantly outperforms Apache in high-traffic scenarios. Its key feature is LSCache, a built-in, server-level page cache that serves pre-compiled pages directly from memory, bypassing PHP and database execution for An event-driven, high-performance web server that significantly outperforms Apache in high-traffic scenarios. Its key feature is LSCache, a built-in, server-level page cache that serves pre-compiled pages directly from memory, bypassing PHP and database execution for subsequent visits and dramatically reducing Time to First Byte (TTFB). | |
Control Panel | cPanel / Plesk / DirectAdmin | Industry-standard control panels providing a comprehensive GUI for tenant self-management. They also offer extensive APIs (cPanel & WHM API) which Mikatsu likely leverages for automated provisioning, billing integration, and service management. | |
Creation Platform | SitePad (Business License) | A static site generator with a drag-and-drop interface. This is a critical architectural choice. By generating static HTML, CSS, and JS files, it eliminates server-side code execution (e.g., PHP) for content delivery. This drastically reduces the server's attack surface, minimizes CPU/memory usage per visitor, and makes the output perfectly suited for distribution via a Content Delivery Network (CDN). | |
Network Security | Hardware Firewall + WAF | A stateful hardware firewall sits at the network edge for perimeter defense. In front of the web servers, a Web Application Firewall (WAF) (e.g., ModSecurity with OWASP Core Rule Set) performs deep packet inspection on HTTP/S traffic to block common application-layer attacks like SQL injection, Cross-Site Scripting (XSS), and RFI. | |
Storage Subsystem | NVMe SSDs in RAID 10 | Storage arrays are built with Non-Volatile Memory Express (NVMe) SSDs, which connect via the PCIe bus, bypassing the SATA controller bottleneck. This provides superior IOPS and lower latency. A RAID 10 configuration is used to provide both data striping for performance and mirroring for redundancy, offering a robust balance for web hosting workloads. |