BluePi Solutions (UK) Ltd • Subsidiary of The Integrity Consortium

Enterprise Data Center Engineering, Security & GCC Strategy

Complete technical & strategic blueprint covering Data Center classifications, deep industry mapping, core hardware & cooling stacks, Uptime Tiers I–IV, Step-by-Step Construction Lifecycle, Physical & Cloud Security, 24/7 Maintenance & Support SLA, Private vs. Cloud, and BluePi Solutions' GCC Service Portfolio & GTM Strategy.

01. Deep-Dive 6 DC Types (Slides 3–4) 02. Single 42U Rack & Defense (Slides 8–9) 03. Structure, Tiers & Build (Slides 10–13) 04. Security & 24/7 SecOps (Slides 14–16) 05. Maintenance, AMC & Support SLA (Slide 17) 06. Cloud, GCC & Marketing (Slides 18–22)

Turnkey UK & GCC

From FS800 Raised Floor & 20kVA+ UPS to All-NVMe SAN & Hyper-V/VMware HA

Tier I → Tier IV

Step-by-step engineering blueprints for every Tier and Data Center type

Gulf Ready

Aligned with KSA Vision 2030, NCA, SAMA, UAE NESA/ADHICS & Qatar NIA

Mission-Critical White Space

Symmetrical Rack Array with Hot/Cold Aisle Containment & Overhead Fiber Raceways

TIA-942 / Tier IV

Core Concept & Strategic Role

What is a Data Center? — The Digital Nerve Center

A specialized, climate-controlled, and physically fortified facility engineered to run mission-critical compute, storage, and network systems 24/7/365 with zero unplanned interruption.

Why an Office Server Closet is NOT a Data Center

Mission-Critical Engineering

Every banking transaction, hospital EHR, ERP workflow, and AI workload depends on continuous infrastructure. Unlike a standard room with split AC and wall power, an engineered Data Center combines anti-static raised flooring (FS800), double-conversion online UPS + generator autonomy, precision sensible-heat InRow/CRAC cooling, clean-agent FM-200/Novec fire suppression, and multi-path carrier fiber.

01 // Compute

Processing Engine

Enterprise x86/ARM rack servers (e.g., HPE ProLiant DL380 Gen12), blade chassis, and AI GPU clusters running hypervisors (Hyper-V, VMware, KVM) and containers.

2.3 kW (Mini DC) → 100+ kW/Rack (AI)

02 // Storage

Data Persistence

Dual-controller All-NVMe SAN arrays (e.g., HPE Alletra MP over 32Gb/64Gb Fibre Channel), NAS clusters, and immutable 3-2-1 backup repositories (RAID-6 + S3 Object Lock).

Sub-Millisecond I/O • RPO ≈ 0

03 // Network

High-Speed Fabric

Redundant Core/Spine-Leaf switching (10G/25G/100G/400G), Active-Passive Next-Gen Firewalls (FortiGate/Palo Alto), WAF load balancers, and dual-ISP demarcation.

Sub-3s HA Failover • Zero Bottleneck

4 Generations of Data Center Evolution 1960s → 2026+

Gen 1: Mainframe

Monolithic Rooms: Siloed mainframes, proprietary cabling, manual tape backups.

Gen 2: Virtualized

42U Racks & Hypervisors: VMware/Hyper-V server consolidation, Fibre Channel SANs, Tier-III Colocation.

Gen 3: Cloud/SDDC

Software-Defined & Hybrid: Hyperscale AWS/Azure/GCP + private cloud HCI with automated API provisioning.

Gen 4: Sovereign/AI

AI Factories & Sovereign Micro-DCs: Direct liquid cooling, 5G edge nodes, and compliant on-prem mini DCs.

4 Engineering KPIs Every Data Center is Judged By

1. PUE (Power Efficiency)

Total Facility Power ÷ IT Load (Target: 1.15–1.45)

2. Uptime SLA ("Nines")

99.9% (HA Mini DC) up to 99.995% (Tier IV)

3. RTO & RPO Resilience

Recovery Time (minutes) & Recovery Point (zero/hours)

4. Thermal Headroom

BTU/hr removal capacity vs. peak server dissipation

Detailed Architectural Taxonomy (Part 1 of 2)

Types of Data Centers — Enterprise, Colocation & Managed

In-depth engineering, commercial, and operational specifications for the three core enterprise facility models.

TYPE 01 // ON-PREMISE 100% Client Owned

Enterprise Data Center (Mini DC to Campus Hall)

A private facility built inside a corporate headquarters (e.g., 16×17 ft Single-Rack HA Mini DC) or a dedicated building owned and operated exclusively by one organization.

Architecture & Subsystems: FS800 antistatic raised floor (300–600mm), dedicated InRow/CRAC precision cooling (25kW–500kW), online double-conversion UPS (20kVA–1MVA), FM-200 clean-agent gas suppression, and private SAN/virtualization cluster.
Ownership Demarcation: Organization owns 100% of civil room, power/cooling plant, 42U racks, servers, storage, and firewalls.
Key Advantages: Absolute physical & legal data sovereignty, zero recurring rack-rent or cloud egress bills, sub-millisecond LAN speed (<0.5ms), tailored hardware.
Trade-offs: Higher upfront CapEx; requires internal IT/facility maintenance or a turnkey partner like BluePi Solutions.

Scale

1 to 200+ Racks

Power Load

3 kW – 2 MW

Build Time

12–16 Wks (Mini)

TYPE 02 // SHARED FACILITY Rent Space, Own IT

Colocation (Colo) Data Center (Retail & Wholesale)

A multi-tenant Tier-III/IV facility operated by a specialist (e.g., Equinix, Khazna UAE, Center3 KSA, Digital Realty) where clients lease locked cabinets, private cages, or whole suites.

Retail vs. Wholesale Colo: Retail Colo rents 1 to 10 racks (10–100 kW); Wholesale Colo leases private halls (500 kW to 10+ MW) to banks and cloud providers.
Ownership Demarcation: Colo provider guarantees building shell, 2N UPS/generators, chilled water HVAC, 6-ring biometric security, and Meet-Me Room (MMR); Client supplies & manages servers, SAN, and firewalls.
Key Advantages: Instant access to certified Tier-III/IV uptime (99.982%+) and rich carrier-neutral ISP cross-connects without building a power plant.
Trade-offs: Monthly kW power + rack rental fees; physical travel or "Remote Hands" fee required for hardware swaps.

Billing

Per Rack + Per kW

Tier Standard

Tier III / Tier IV

Go-Live

4–8 Weeks

TYPE 03 // TURNKEY OPS 100% Provider Managed

Managed Services & Hosted Private Cloud DC

A single-tenant dedicated infrastructure model where the service provider owns both the facility AND the physical servers/SAN, operating them under a strict Managed SLA.

What is Included: Dedicated bare-metal hosts, hardware replacement within 2–4 hours, hypervisor management, OS patching, firewall administration, 24/7 NOC monitoring, and daily Veeam backup verification.
Ownership Demarcation: Provider owns facility + hardware + hypervisor lifecycle; Client owns only its applications, databases, and business data.
Key Advantages: Converts massive hardware CapEx into predictable monthly OpEx; ideal for organizations lacking a large 24/7 internal infrastructure engineering team.
Trade-offs: Higher 5-year recurring cost than self-owned hardware; standardized hardware catalog.

Financials

100% Monthly OpEx

Hardware Risk

Zero (Provider SLA)

Go-Live

1–3 Weeks

Detailed Architectural Taxonomy (Part 2 of 2)

Types of Data Centers — Hyperscale, Edge, Modular & Full Matrix

Comparing massive cloud campuses, decentralized low-latency edge micro-nodes, and containerized rapid-deployment facilities.

TYPE 04 // HYPERSCALE & AI 50 MW – 1,000+ MW

Cloud & Hyperscale Data Center

Scale: 50,000 to 500,000+ servers across massive campuses (AWS, Azure, GCP, Meta).
Engineering: Custom open-compute racks, 400G/800G optical fabric, direct-to-chip liquid cooling for 40–100kW AI GPU racks, industry-best PUE (1.08–1.15).
Consumption: Consumed by enterprises as virtualized IaaS/PaaS/SaaS over the internet or private Direct Connect.

TYPE 05 // DECENTRALIZED EDGE <1 ms – 5 ms Latency

Edge & Micro Data Center

Scale: 1 to 10 ruggedized racks (5 kW–100 kW) placed at 5G telecom base stations, hospital floors, oil rigs, or smart city intersections.
Engineering: Soundproof, dust-sealed IP55 cabinets with integrated UPS, in-cabinet DX cooling, and autonomous lights-out remote management.
Role: Processes high-volume video/IoT telemetry locally before syncing summaries to central cloud.

TYPE 06 // PREFABRICATED 20ft / 40ft ISO Modules

Modular / Containerized Data Center

Scale: Factory-assembled, weatherproof steel modules housing 6 to 20 racks (50 kW–500 kW per module).
Engineering: Pre-wired at factory with racks, InRow cooling, UPS, and FM-200 gas; craned onto a concrete pad and energized in 6–10 weeks (vs. 18 months civil build).
Role: Ideal for GCC oilfields, NEOM/mega-project construction sites, military bases, and instant Disaster Recovery.

Data Center Type	Typical Rack Count	Power Capacity	Density (kW/Rack)	Typical PUE	CapEx vs. OpEx	Primary Owner of Hardware
1. Enterprise On-Prem (Mini / Core)	1 – 150 Racks	2.3 kW – 1.5 MW	2.5 – 12 kW	1.35 – 1.60	High CapEx / Low OpEx	Client (100% Facility + IT)
2. Colocation (Retail / Wholesale)	1 Cage – 1,000+ Racks	10 kW – 20 MW	5 – 30 kW	1.20 – 1.35	Hybrid CapEx + OpEx	Colo (Facility) / Client (IT)
3. Managed Data Center	1 – 50 Dedicated Racks	5 kW – 500 kW	5 – 15 kW	1.25 – 1.40	100% Predictable OpEx	Managed Provider (100%)
4. Cloud / Hyperscale	10,000 – 100,000+ Racks	50 MW – 1,000 MW	15 – 100+ kW (AI)	1.08 – 1.18	100% Elastic OpEx	Hyperscaler (AWS/Azure/GCP)
5. Edge Micro-Data Center	1 – 10 Racks per site	3 kW – 100 kW	3 – 20 kW	1.30 – 1.50	Modular CapEx	Telco / Industrial Enterprise
6. Prefabricated Modular Container	6 – 24 Racks / Container	50 kW – 1 MW	8 – 25 kW	1.20 – 1.35	Fast-Track CapEx	Client or Turnkey Lease

Industry-by-Industry Architecture Alignment (Part 1 of 2)

Who Uses Which Data Center? — 6 Core Industry Profiles

Detailed mapping of how different industries combine On-Premise, Colocation, Edge, and Cloud based on regulatory law, latency physics, and workload economics.

SECTOR 01 // SOVEREIGN & DEFENSE Tier IV On-Prem

Government, Defense & Central Banks

Users: Ministries of Interior/Defense, National ID Registries, Central Banks (SAMA, CBUAE), Tax Authorities, Intelligence.

DC Used: 100% Sovereign Enterprise Private DC (On-Premise) + Physically Air-Gapped Disaster Recovery DC (>200 km apart).
Why: National security laws strictly forbid citizen biometrics, military telemetry, or sovereign monetary ledgers from residing on foreign-owned public clouds.

SECTOR 02 // BFSI & HEALTHCARE Tier III/IV Hybrid

Commercial Banks, FinTech & Hospitals

Users: Retail/Islamic Banks, Payment Switches, Capital Markets, Large Hospital Networks (HIS/PACS imaging).

DC Used: On-Premise / Tier-IV Colocation Core + Local Sovereign Cloud (Hybrid Model).
Why: Core Banking (Temenos/Finacle), ATM/POS card switches (PCI-DSS), and 4K MRI hospital PACS require sub-millisecond local SAN speed and strict data residency (SAMA/ADHICS), while mobile banking apps scale on cloud.

SECTOR 03 // CORPORATE & SME Turnkey Mini DC + Cloud

Corporate HQs, EPC, Real Estate & Trading

Users: 50–500 employee HQs, Family Conglomerates, Logistics/Shipping, Construction/EPC, Manufacturing plants.

DC Used: On-Premise Single-Rack High-Availability Mini DC (42U) or Retail Colocation + Cloud Off-Site Backup (3-2-1).
Why: Runs Active Directory, ERP (SAP/Oracle/Odoo), file shares, CCTV NVR, and IP-PBX locally with zero LAN lag, backed up immutably to AWS S3/Wasabi.

SECTOR 04 // ENERGY, TELCO & IOT Modular + Edge DC

Oil & Gas, 5G Telecom, Ports & Smart Cities

Users: Aramco/ADNOC oilfields, 5G operators (STC, e&, du, Ooredoo), Automated Seaports, Smart Traffic & Utilities.

DC Used: Prefabricated Containerized Modular DCs + Distributed Edge Micro-DCs.
Why: Offshore rigs and desert drilling sites have high-latency satellite backhaul—seismic data and SCADA safety controls must be processed on-site inside ruggedized containers.

SECTOR 05 // DIGITAL NATIVES Multi-AZ Public Cloud

E-Commerce, SaaS, Ride-Hailing & Media

Users: Online Marketplaces, Delivery Super-Apps, EdTech, OTT Video Streaming, High-Growth Startups.

DC Used: Public Cloud / Multi-Cloud (AWS, Azure, GCP) + Global Edge CDN (Cloudflare/Akamai).
Why: Traffic surges 10x–50x during Ramadan sales or live sports events; Kubernetes auto-scaling spins up 500 nodes in minutes and scales down overnight to save cost.

SECTOR 06 // AI & RESEARCH Liquid-Cooled Hyperscale

AI Labs, Sovereign LLMs & Supercomputing

Users: National AI Programs (G42,SDAIA/HUMAIN), Universities, Autonomous Driving & Genomics Institutes.

DC Used: Hyperscale AI Data Centers & Wholesale High-Density Colocation (50–100 kW/rack).
Why: NVIDIA H100/B200 GPU clusters require InfiniBand/RoCEv2 lossless networking and direct-to-chip liquid cooling that standard air-cooled halls cannot support.

Quantitative Selection Matrix (Part 2 of 2)

Matching Organization Size, Compliance & Workload to DC Type

How to determine the exact Data Center Type and Uptime Tier based on employee count, VM scale, regulatory law, and RPO/RTO tolerances.

Organization Profile & Size	Typical IT Workload	Recommended DC Model	Target Tier & Redundancy	RTO / RPO Target	Governing Compliance / Standard
SME / Corporate HQ (50 – 250 Users \| 10–30 VMs)	Active Directory, ERP, File/Print, IP-PBX (70+ phones), Local DB	On-Premise 42U HA Mini DC + S3 Cloud Backup	Tier-II Facility + N+1 IT HA (20kVA UPS + 25kW InRow)	RTO: <15 mins RPO: ≤1–24 hrs	ISO 27001, Local Tax/Data Privacy (PDPL / GDPR)
Mid-to-Large Enterprise (250 – 2,500 Users \| 50–300 VMs)	SAP S/4HANA, Oracle DB, CRM, VDI, Corporate Intranet, BI	Primary On-Prem / Colo (3–10 Racks) + Cloud DR	Tier III (99.982%) Dual PDU (32A), 2N UPS, Generator	RTO: <5 mins RPO: <15 mins	ISO 27001, SOC 2 Type II, UAE NESA / KSA NCA ECC
Commercial Bank / FinTech / Hospital (1,000 – 10,000+ Users)	Core Banking, ATM/Card Switch, EHR/PACS, Real-Time Fraud AI	Dual-Site Active-Active (Primary + Far DR) + Hybrid Cloud	Tier III+ / Tier IV (99.995%) 2N Active-Active Power/Cooling	RTO: ~0 (Auto) RPO: 0 (Sync SAN)	SAMA CSF, CBUAE, PCI-DSS v4.0, ADHICS, HIPAA
Industrial / Oil & Gas / Port (Remote Sites / 1,000+ Sensors)	SCADA/ICS, Digital Twin, AI Video Analytics, Local Historians	Modular Container DC / Rugged Edge + Central Cloud	Tier II / III Modular (Dual DX Cooling + Generator)	RTO: <1 min RPO: Real-Time	IEC 62443 (OT Security), Aramco CCC, ADNOC Standards
SaaS / E-Commerce / Digital Platform (10k – 10M+ External Users)	Kubernetes Microservices, NoSQL, Global CDN, Elastic Web	Multi-AZ Public Cloud (AWS / Azure / GCP)	3-AZ Cloud Region (99.99% Multi-AZ SLA)	RTO: Seconds RPO: Seconds	SOC 2 Type II, ISO 27017/27018, PCI-DSS, GDPR

Real-World Reference Benchmark: BluePi Solutions' Turnkey 42U Corporate Mini Data Center

Proven Production Architecture

ROOM & RACK FOOTPRINT

16×17 ft (272 sq.ft) room, 300mm FS800 floor (800 kg/m²), single 42U rack (37U active / 5U buffer, ~555 kg).

COMPUTE & ALL-NVMe SAN

2× HPE DL380 Gen12 hosts (64-core Xeon, 512GB DDR5) + HPE Alletra MP All-NVMe SAN (32Gb FC active-active).

POWER & THERMAL HEADROOM

2.3 kW typical / 5.0 kW peak IT load backed by APC 20kVA UPS (~50m fresh / 38m aged runtime) & 25.1kW InRow PAC.

NETWORK, SECURITY & DR

2× FortiGate FG-201G HA, Cisco C9300X stacks, FM-200 fire gas, and Veeam 3-2-1 RAID-6 + S3 Object Lock.

Inside the Data Center

Core Components — IT Hardware & Facility Subsystems

A data center is an integrated system divided into the IT Stack (Compute, Storage, Network) and the Facility Stack (Power, Cooling, Management).

1. Compute Infrastructure IT Stack

The processing engines mounted inside standardized 19-inch, 42U–48U server racks.

Enterprise Rack Servers (1U–4U): Dual-socket hosts (e.g., HPE ProLiant DL380 Gen12) with redundant hot-swap PSUs.
Blade & Composable Systems: High-density modular enclosures sharing power and fabric.
AI / GPU Clusters: High-wattage accelerator nodes for LLM training and inference.
Hypervisor Cluster: Windows Server Datacenter (Hyper-V Failover) or VMware vSphere HA.

2. Enterprise Storage IT Stack

Fault-tolerant storage arrays protecting production databases, VMs, and immutable backups.

All-NVMe SAN (Fibre Channel): Dual active-active controllers (e.g., HPE Alletra MP B10100) connected via 32Gb/64Gb MPIO FC.
Dedicated Backup Server: Standalone host (RAID-6 SAS repository) isolated from main SAN fault domain.
NAS & Object Storage: Unstructured file sharing and S3 Object Lock WORM storage.
3-2-1-1-0 Backup Rule: 3 copies, 2 media types, 1 off-site cloud, 1 immutable/air-gapped, 0 errors.

3. Network & Security Fabric IT Stack

Non-blocking optical and copper switching fabric with hardware perimeter defense.

Core & Server Switches: Stacked 25G/10G/100G switches (e.g., Cisco Catalyst C9300X with StackWise-1T & StackPower).
Next-Gen Firewalls (NGFW): Active-Passive HA cluster (e.g., 2× FortiGate FG-201G) with IPS, AV, and WAF (KEMP LoadMaster).
WAN & OOB Switches: Dual-ISP WAN switches + dedicated Out-of-Band management switch (iLO 7, SAN, UPS, KVM).
Structured Cabling: OM4/OS2 fiber LIUs and Cat6A shielded patch fields.

4. Electrical & Power Chain Facility Stack

Conditioned, uninterruptible power delivery preventing voltage sags, spikes, or blackouts.

Online Double-Conversion UPS: e.g., APC Smart-UPS RT 20kVA–500kVA with extended battery packs (IEEE 1188 aging derating accounted for).
Dual Smart Rack PDUs: Vertical 0U metered/switched PDUs (PDU-A & PDU-B; 16A/32A rated) feeding dual-corded server PSUs.
Standby Generator & ATS: Diesel generator with Automatic Transfer Switch (<60s cutover) and external manual maintenance bypass.
Clean Technical Earth: Dedicated copper earth pit (<1 Ohm resistance).

5. Precision Cooling & Floor Facility Stack

Removes continuous sensible heat (BTU/hr) while maintaining 18°C–27°C and 40–60% RH.

InRow / CRAC Precision AC: Close-coupled 25.1kW+ InRow PAC or downflow CRAC designed for 24/7 sensible heat removal (unlike comfort wall ACs).
Antistatic Raised Floor: Heavy-duty FS800 600×600mm tiles on 300mm–600mm pedestals (800 kg/m² uniform load) with entrance ramp.
Thermal Isolation: 1U/2U rack blanking panels & hot/cold aisle containment preventing exhaust air recirculation.

6. Fire Safety, OOB & DCIM Facility Stack

Automated hazard suppression and lights-out remote management across the facility.

Clean-Agent Fire Suppression: Automated FM-200 (HFC-227ea) or Novec 1230 total flooding gas system interlocked with building FACP & EPO switch.
Rack Crash-Cart KVM: 1U slide-out 18.5" LCD KVM console (e.g., ATEN CL3708NX) + iLO/iDRAC out-of-band BIOS access.
DCIM & Environmental Telemetry: Real-time temperature, humidity, water-leak, and UPS dry-contact alerting.

Unit-by-Unit Hardware Elevation, Wattage & BTU Breakdown (U1 to U42 + 0U Rear)

Inside a Single 42U Server Rack — U1 to U42 Elevation Blueprint

Complete unit-by-unit hardware map from bottom (258.5kg UPS/Battery) to top (WAN & NGFW Defense)—37U Occupied (88.1%) / 5U Free Buffer (11.9%).

Single 42U Enterprise Server Rack Cutaway Anatomy

42U RACK ELEVATION ANATOMY

U1–U11 Power | U12–U19 SAN & Compute | U21 KVM | U24–U30 Voice & Patch | U31–U42 Network & Defense

Verified Rack Weight, Power & Thermal Physics FS800 & 20kVA Verified

STATIC WEIGHT

~555 kg Total

UPS: 142.5kg | Batt: 116kg | IT/SAN: 157kg | Rack: 140kg

POWER & HEAT

2,304W / 5,041W

~7,860 BTU/hr core (~50.2m fresh / ~38.6m aged UPS)

0U & ADJACENT

A/B PDU + PAC

2× ATEN PE8216G + 25.1kW InRow AC + FM-200

U-Slot	Exact Hardware Model & Qty	Typ / Peak W	Subsystem Role, Ports & Redundancy Spec
U41–U42	2× FortiSwitch FS-108F WAN Switches (2U)	40W / 48W	Dual ISP fiber demarcation splitting ISP-A & ISP-B into both NGFWs
U39–U40	2× FortiGate FG-201G NGFW HA Pair (2U)	104W / 228W	Active-Passive Perimeter Defense (<3s RTO, 3-Yr FortiGuard UTP, IPS, SSL)
U37–U38	2× Cisco Catalyst C9300X-24Y-M Core (2U)	300W / 700W	24× 25G/10G SFP28 Core Stack (StackWise-1T 1Tbps + StackPower)
U35–U36	2× Cisco Catalyst C9300X-12Y-M Server (2U)	240W / 700W	12× 25G/10G SFP28 Server Farm Stack (LACP dual-homed to all hosts)
U34	1× FortiSwitch FS-124F OOB Switch (1U)	25W / 30W	Air-gapped Out-of-Band mgmt network (iLO 7, Alletra SAN, KVM, PDU, UPS)
U31–U33	3× Cisco Catalyst C9200L-24PXG PoE+ (3U)	430W / 1,350W	72× PoE+ mGig ports + 10G SFP+ uplinks for 70 users, Wi-Fi & IP phones
U26–U30	OM4 Fiber LIU + Cat6A Patch + 1U Blank (5U)	0W (Passive)	4U Structured cabling field (26× OM4 LC + Cat6A STP) + U30 thermal blank
U24–U25	1× Yeastar P560 IP-PBX + 1U Blank (2U)	25W / 45W	Corporate voice controller (E1 PRI / SIP trunk for 70 Yealink IP phones)
U20–U23	ATEN CL3708NX 18.5" LCD KVM + Buffers (4U)	25W / 35W	8-Port slide-out Full-HD KVM crash-cart (U21) + 3U thermal/cable buffers
U16–U19	2× HPE ProLiant DL380 Gen12 Hosts (4U)	840W / 1,600W	Host 1 & 2: 2× Xeon 6730P (64C), 512GB DDR5, 2× 32Gb FC, 4× 10G SFP+
U14–U15	1× HPE ProLiant DL380 Gen12 Backup (2U)	280W / 800W	Physical Veeam Host: Xeon 6515P, 6× 12TB SAS (72TB Raw / ~48TB RAID-6)
U12–U13	1× HPE Alletra MP B10100 All-NVMe SAN (2U)	450W / 900W	Dual Active-Active Nodes, 12× 1.92TB NVMe (15TB licensed), 32Gb FC MPIO
U1–U11	APC 20kVA UPS (7U) + 192V Battery (4U) (11U)	20kW Cap.	SRTG20KXLI (142.5kg) + SRTG192XLBP2 (116kg) + Rear 0U Dual ATEN PDUs

Deep-Dive Functional Breakdown: Defense System, Switching, Compute, NVMe SAN, Backup & Power

Inside a Single 42U Rack — Defense System & All 6 Subsystems

Detailed architectural breakdown of every subsystem inside the 42U rack—from the 7-Layer Rack Defense System down to Compute, All-NVMe SAN, 3-2-1 Backup, and Power/Cooling.

1. Multi-Layer Defense System

U41–42, U39–40, U34

7 concentric cyber, hardware, and physical defense layers inside a single rack:

WAN Demarcation (U41–42): 2× FortiSwitch FS-108F isolate dual ISP fiber links before hitting internal zones.
NGFW HA Perimeter (U39–40): 2× FortiGate FG-201G (Active-Passive <3s failover) with 3-Yr FortiGuard UTP (IPS, Anti-Botnet, Deep SSL Inspection, IPsec VPN).
Virtual WAF Layer: 2× KEMP LoadMaster WAF VMs block OWASP Top-10 (SQLi/XSS) & L7 DDoS attacks.
Air-Gapped OOB & Root-of-Trust: FortiSwitch FS-124F (U34) isolates iLO 7/SAN/UPS; HPE Silicon Root of Trust + TPM 2.0 + 10× VM EDR agents + Biometric Rack Handle Lock.

7-Checkpoint Zero-Trust Packet Inspection

2. Switching & Cabling Fabric

U31–U38 & U26–U29

Non-blocking 25G/10G East-West backbone with hardware stacking and PoE+ access:

Core Backbone (U37–38): 2× Cisco Catalyst C9300X-24Y-M (24× 25G/10G SFP28) bonded via StackWise-1T (1 Tbps) & StackPower shared PSU bus.
Server Aggregation (U35–36): 2× Cisco C9300X-12Y-M connecting dual-homed 10G SFP+ NICs from all 3 HPE servers via LACP port-channels.
PoE+ Access & Voice (U31–33 & U24): 3× Cisco C9200L-24PXG (72× PoE+ ports) + Yeastar P560 IP-PBX (E1 PRI/SIP) for 70 users & IP phones.
Patch Field (U26–29): 26× OM4 multimode fiber LC ports + Cat6A shielded patch panels.

1 Tbps StackWise-1T • 802.1X / VLAN Micro-Seg

3. HA Compute Cluster

U16–U19 (128 Cores)

Twin enterprise virtualization hosts engineered for 100% N+1 workload survival:

2× HPE ProLiant DL380 Gen12 (U16–17 & U18–19): Each node packs 2× Intel Xeon 6730P (64 Cores/node = 128 Cores total) and 512 GB DDR5 RAM (1 TB total).
Boot & I/O Fabrics: 2× 960GB NVMe M.2 (RAID-1 OS boot), 2× 32Gb Fibre Channel HBA ports, 4× 10Gb SFP+ network ports, and 2× 1000W Titanium hot-plug PSUs.
Hypervisor & Licensing: 8× 16-Core Windows Server Datacenter packs + 70 User CALs running Hyper-V Failover Clustering & Live Migration.
Production Workloads: 10 Core VMs (AD DS, DNS, ERP, SQL DB, File Server, 2× KEMP WAF).

128 Xeon Cores • 1 TB DDR5 • Live Migration

4. All-NVMe SAN Storage

U12–U13 (32Gb FC)

Tier-1 mission-critical block storage with zero single point of failure:

HPE Alletra Storage MP B10100 (2U): Powered by HPE ArcusOS with Dual Active-Active Controller Nodes for 100% uptime guarantee.
All-NVMe Flash Media: Populated with 12× 1.92TB NVMe SSDs (23.04 TB Raw / 15 TB Usable Licensed) delivering sub-millisecond database I/O.
Direct-Attach 32Gb Fibre Channel MPIO: Both compute hosts and the backup server connect directly via redundant 32Gb FC links (including 4× host-side 32Gb FC transceivers absorbed turnkey by BluePi).
Data Protection: Hardware AES-256 encryption at rest, instant snapshots, and hardware RAID parity.

Dual Active-Active Controllers • <0.5ms Latency

5. 3-2-1 Backup & KVM Console

U14–U15 & U21

Physically isolated ransomware recovery vault and direct in-rack crash-cart control:

Standalone Physical Backup Server (U14–15): Dedicated HPE ProLiant DL380 Gen12 (Xeon 6515P, 64GB RAM) isolated from the production Hyper-V cluster.
72TB Raw RAID-6 Vault: Populated with 6× 12TB SAS HDDs (~48 TB Usable RAID-6) + dual 32Gb FC SAN direct-read for LAN-free Veeam backups.
Cloud S3 Immutability: Veeam 10-VM Universal License replicates encrypted backups to off-site Wasabi/AWS S3 Object Lock (WORM).
Local KVM Console (U21): ATEN CL3708NX 8-port slide-out 18.5" Full-HD LCD monitor, keyboard & touchpad drawer.

RPO ≤ 24h • Instant VM Recovery in Minutes

6. Power, PDU & Cooling

U1–U11, 0U & Row

Heavy-duty power conditioning, A/B feed redundancy, and close-coupled cooling:

20kVA Online UPS + Battery (U1–U11): APC Smart-UPS RT 20kVA (SRTG20KXLI, 142.5kg) + 192V Extended Battery (SRTG192XLBP2, 116kg = 258.5kg). Delivers ~50.2 min fresh / ~38.6 min aged runtime at 2.3 kW typical load.
Dual 0U Smart PDUs (Rear Sides): 2× ATEN PE8216G 16A (3.68kW max / 2.94kW cont.; 32A upgrade optional) splitting load ~1.15 kW per side.
Bypass & Generator: External manual maintenance bypass panel + facility generator (<60s ATS).
Thermal & Fire Envelope: Adjacent 25.1kW InRow PAC + FM-200 clean-agent gas + 300mm FS800 floor.

0ms Double-Conversion • ~50m Battery Autonomy

Architectural Anatomy: Room Zones (Ceiling, White Space, Underfloor) & Rack Frame Structure

Data Center Structure & Physical Rack Anatomy

How the data center room is structured across 3 vertical layers (Overhead, Hall, Underfloor) and the 10 structural parts engineered inside every 42U Rack Cabinet.

3D Isometric Data Center Cutaway Diagram

3D FACILITY & RACK STRUCTURE CUTAWAY

White Space (Racks, InRow AC, FS800 Floor) | Grey Space (UPS, Bypass, Generator) | MMR Vault

TIA-942 Layout

3-Layer Vertical Room Structure (16×17 ft / Enterprise Hall)

Ceiling → Floor

1. CEILING / OVERHEAD RACEWAY ZONE

Segregated OM4/OS2 Fiber Runner Tray & Power Basket Tray, VESDA Laser Smoke Pipe, FM-200 Discharge Nozzles, & 4K AI IR Dome CCTV.

2. WHITE SPACE HALL vs. GREY SPACE PLANT

Front Cold Aisle (18–22°C) & Rear Hot Aisle, adjacent 25.1kW InRow PAC, FM-200 Cylinder + FACP/EPO, and wall-mounted Manual UPS Bypass Panel.

3. UNDERFLOOR PLENUM (300mm–600mm FS800)

FS800 Antistatic Tiles (800 kg/m²) + Entrance Ramp; beneath floor: copper AC refrigerant/drain lines, armored power feeds, <1Ω Earth Busbar & Water-Leak Rope.

Physical Structure of a 42U Rack Cabinet (600×1070mm Frame Anatomy)

10 Structural Parts

Beyond the servers and switches (covered in Slide 8), the physical 42U Server Rack Enclosure itself consists of 10 engineered structural subsystems:

01. 19" EIA-310 SEISMIC FRAME

Welded steel frame with 4 adjustable square-hole (cage-nut) vertical rails numbered U1 to U42 (rated up to 1,500 kg static load).

02. PERFORATED MESH DOORS (80%+)

Curved hexagonal-mesh front door (cold intake) + split perforated rear French doors (hot exhaust) for unobstructed front-to-back airflow.

03. SMART HANDLE LOCKS & SENSORS

Electronic biometric/RFID swing-handle locks + magnetic door-contact & temperature/humidity probes wired to DCIM/SIEM.

04. REAR 0U VERTICAL PDU CHANNELS

Dedicated rear-corner toolless brackets holding PDU-A (Left) & PDU-B (Right) (2× ATEN PE8216G) consuming zero horizontal U-space.

05. TOP BRUSH ENTRY & FAN TRAY

Dust-sealed brush gland openings on the rack roof for overhead fiber/Cat6A drops + top-mounted thermal exhaust fan tray.

06. SLIDING RAILS & CABLE ARMS (CMA)

Ball-bearing telescopic rail kits + articulated rear Cable Management Arms allowing live servers to slide out without unplugging cords.

07. VERTICAL & HORIZONTAL FINGER DUCTS

Vertical D-rings & 1U/2U horizontal cable managers (U26–U29) maintaining strict bend-radius separation between OM4 fiber and Cat6A copper.

08. 1U/2U THERMAL BLANKING PANELS

Snap-in steel blanking plates installed across all 5U unused slots (U20, U22–23, U25, U30) to block hot exhaust air recirculation.

09. COPPER GROUNDING BUSBAR (<1Ω)

Full-height tinned copper grounding bar bonding rack doors, side panels, and server chassis directly to the underfloor clean earth grid.

10. LEVELING FEET, CASTERS & PLINTH

Heavy-duty rolling casters, 4 load-leveling feet, baying brackets (linking to InRow AC), and seismic anchor bolts into the FS800 floor.

Uptime Institute Global Benchmark

Data Center Structure — Tier I, II, III & IV Standards

The Uptime Institute classifies data center structural resilience into four progressive Tiers based on redundancy, concurrent maintainability, and fault tolerance.

TIER I 99.671%

Basic Capacity

Single non-redundant path for power and cooling. Planned maintenance requires complete IT shutdown.

Redundancy: N (None)
Downtime/Yr: ≤ 28.8 Hours
Staffing: Business hours only

TIER II 99.741%

Redundant Components

Single distribution path with redundant components (UPS, generator, chillers). Line maintenance still requires shutdown.

Redundancy: N+1 Components
Downtime/Yr: ≤ 22.0 Hours
Use Case: SME / Corporate Mini DCs

TIER III 99.982%

Concurrently Maintainable

Dual power/cooling distribution paths (1 active, 1 alternate). Any part can be serviced with zero IT downtime.

Redundancy: N+1 Equipment / Dual Path
Downtime/Yr: ≤ 1.6 Hours (95 mins)
Use Case: Enterprise Colo, Banks, SaaS

TIER IV 99.995%

Fault Tolerant

Two independent, physically fire-compartmentalized active paths. Survives any single unplanned failure automatically.

Redundancy: 2N or 2(N+1) Active-Active
Downtime/Yr: ≤ 26.3 Minutes
Use Case: Central Banks, Defense, Core Cloud

Internal Network Structure: Legacy 3-Tier vs. Modern Spine-Leaf Clos Fabric

East-West Traffic Optimization

Legacy 3-Tier Hierarchy (Core → Aggregation → Access)

Built for North-South client-server traffic. Spanning Tree Protocol (STP) blocks redundant links and creates variable multi-hop latency during VM live migration or storage replication.

Modern Spine-Leaf / StackWise-1T Fabric (Full Mesh)

Every Top-of-Rack Leaf switch connects to every Spine switch via active-active Layer-3 ECMP or 1-Tbps hardware stacking—delivering deterministic microsecond East-West latency.

Turnkey Construction & Integration Methodology

How to Build a Data Center Step-by-Step (7-Phase Lifecycle)

From bare concrete slab to production workload cutover—the exact sequential engineering workflow used to construct a reliable Data Center.

Progressive Step-by-Step Data Center Construction Cutaway

PROGRESSIVE TURNKEY BUILD

Left-to-Right: Concrete Subfloor & FS800 Pedestals → InRow Cooling & UPS → Cabled 42U Racks & FM-200

16-Week Fast-Track Implementation Gantt (Mini/Enterprise DC) Turnkey Schedule

WEEKS 1 – 4

Design & Civil Prep: Load study, HLD/LLD sign-off, hardware procurement, room sealing, subfloor leveling.

WEEKS 5 – 8

Floor, Power & HVAC: 300mm FS800 floor & ramp, clean earth (<1Ω), InRow PAC piping, FM-200 gas nozzles.

WEEKS 9 – 12

Rack, SAN & Network: 42U rack, 20kVA UPS + 32A PDUs, DL380 hosts, Alletra NVMe SAN, C9300X & FG-201G HA.

WEEKS 13 – 16

Cluster, Backup & UAT: Hyper-V/VMware HA, Veeam 3-2-1 S3 immutability, 8-point failover UAT & handover.

The 7 Sequential Steps to Create a Production Data Center

End-to-End Blueprint

STEP 01 // SIZE

Workload Audit & Capacity Sizing: Calculate user count, VM CPU/RAM/IOPS, electrical peak/typical load (W), heat dissipation (BTU/hr), and target Uptime Tier (Tier I–IV).

STEP 02 // CIVIL

Room Shell & Raised Access Floor: Construct 2-hr fire-rated walls, vapor barrier, dust-proof slab, and install FS800 600×600mm antistatic raised floor (300–600mm height) + ramp.

STEP 03 // POWER

Electrical & UPS Chain: Provision dedicated sub-DB, clean technical earth (<1Ω), generator + ATS (<60s), external maintenance bypass, Online UPS, and dual A/B Rack PDUs.

STEP 04 // HVAC

Precision Cooling & Fire Suppression: Install InRow PAC / CRAC units with outdoor condensers, hot/cold containment, VESDA smoke sensors, and FM-200 / Novec 1230 gas cylinders.

STEP 05 // CABLE

Racks, Structured Cabling & Security: Anchor 42U racks, dress overhead/underfloor OM4 fiber & Cat6A patch panels, and wire biometric access control, CCTV NVR & EPO relays.

STEP 06 // IT HA

Mount Active IT & Storage Fabric: Rack WAN switches, HA NGFW firewalls, Core/Server switch stacks, dual-controller All-NVMe SAN (32Gb FC MPIO), compute hosts & KVM.

STEP 07 // UAT

Cluster Provisioning, DR & IST/UAT Drills: Build Hyper-V/VMware cluster, configure Veeam 3-2-1 S3 immutable backups, execute live power/network failover drills & sign off As-Built docs.

Tier-by-Tier & Type-by-Type Engineering Guide

How Each Data Center Tier (I–IV) & Type is Engineered

What specific infrastructure must be added at each level to build a Tier I, Tier II, Tier III, or Tier IV facility—plus build steps by DC Type.

Target Tier	Step 1: Power & Utility Build	Step 2: UPS & Generator Build	Step 3: Cooling & Pipework Build	Step 4: IT, Network & Compartments
BUILD TIER I Basic (N)	Pull 1 utility feed to a single main distribution board (MDB) and single rack PDU.	Install 1 UPS module (N) for short battery ride-through; optional single standby generator.	Install 1 CRAC / DX cooling unit (N) with single refrigerant/chilled line.	Single ISP router, single switch, standalone server. Must shut down IT for any maintenance.
BUILD TIER II Redundant (N+1)	Single utility path + external manual bypass panel + Dual Rack PDUs (A & B) inside 42U rack.	Install Online UPS (N+1 modules) + 1 Diesel Generator with ATS (<60s) & 24hr fuel tank.	Install InRow/CRAC units in N+1 (e.g., 25kW cooling for 5kW load) or backup cooling unit.	BluePi HA Model: Dual-PSU servers, 32Gb FC MPIO SAN, Stacked Core Switches, Active-Passive Firewalls.
BUILD TIER III Maintainable	Build 2 completely independent power paths (Active Path A + Alternate Path B) from sub-station to rack.	Install Dual UPS systems (N+1 or 2N) + N+1 Generators with 72hr fuel & isolation breakers.	Build Dual-pipe chilled water loops with isolation valves so any chiller/pipe is serviced live.	Dual Meet-Me Rooms (North/South fiber entry), Spine-Leaf network. Zero downtime during maintenance.
BUILD TIER IV Fault Tolerant (2N)	Pull 2 independent HV utility grids from separate substations into fire-separated switchrooms.	Build 2N or 2(N+1) Active-Active UPS & Generator plants in separate 2-hr fire-rated zones.	Build 2N independent chiller plants & continuous cooling batteries (ride-through thermal storage).	Full physical compartmentalization of every A & B path + autonomous fault-isolation controls.

HOW TO BUILD // ON-PREM DC

1. Enterprise / Mini DC

1) Select ground/mid-floor room away from wet pipes → 2) Fit FS800 floor, 25kW InRow AC & FM-200 → 3) Wire 20kVA UPS & 32A PDUs → 4) Deploy 42U rack, HPE hosts, NVMe SAN, FortiGate HA & Veeam.

HOW TO BUILD // COLOCATION

2. Colocation Cage Build

1) Lease kW & cage in Tier-III/IV Colo (Equinix/Khazna/Center3) → 2) Order dual A+B 32A whips & MMR ISP cross-connects → 3) Ship & rack servers/SAN/firewalls → 4) Establish IPsec/Direct Connect to HQ.

HOW TO BUILD // MODULAR EDGE

3. Container / Edge DC

1) Pour outdoor concrete pad & generator fuel base → 2) Factory-integrate 20ft/40ft ISO container with racks, DX cooling, UPS & fire gas → 3) Crane container onto pad → 4) Plug HV power & fiber/5G uplink.

HOW TO BUILD // CLOUD DC

4. Virtual Cloud DC (SDDC)

1) Architect Multi-AZ Landing Zone in AWS/Azure/GCP using Terraform → 2) Configure Hub-Spoke VPCs, Transit Gateway & NGFW → 3) Provision KMS encryption, IAM Zero Trust & EKS/VM auto-scaling groups.

Defense-in-Depth Facility Protection

Physical Security & Fire/Environmental Safety

Data centers enforce 6 concentric physical security rings alongside automated clean-agent fire suppression that extinguishes fires without damaging electronics.

Biometric Mantrap and Security Operations Center

ZERO-TAILGATING ACCESS CONTROL

Interlocking Biometric Mantrap Portal (Iris + Palm Vein MFA) & 24/7 SOC Wall

Fire & Environmental Safety Systems Zero Water Damage

VESDA (Very Early Smoke Detection Apparatus): Laser air-sampling sensors detect microscopic incipient smoke particles minutes before visible flame.
Clean Agent Gas Suppression (FM-200 / Novec 1230 / Inergen): Total flooding gas extinguishes fire in <10s by interrupting combustion with zero water or residue damage to servers.
FACP & EPO Interlock: Interlocked with Building Fire Alarm Control Panel, HVAC damper shutoff (to contain gas), and guarded Emergency Power Off (EPO) button.
Under-Floor Leak & Seismic Protection: Hydro-sensing cables under raised floor detect condenser drips; seismic rack anchoring protects against tremors.

The 6 Concentric Rings of Data Center Physical Security

ISO 27001 / NCA / SAMA Physical Controls

RING 1 // SITE

Perimeter Defense: Anti-climb electrified fencing, K12 crash-rated hydraulic bollards, vehicle traps, and blast-standoff buffer zones.

RING 2 // YARD

Surveillance & Patrol: 360° AI thermal/IR CCTV dome cameras with 90-day NVR retention, motion radar, and 24/7 security guards.

RING 3 // LOBBY

Identity Vetting: Pre-registered work permit check, national ID/passport verification, metal detection, and X-ray bag screening.

RING 4 // PORTAL

Single-Person Mantrap: Interlocking bulletproof glass airlocks with weight sensors (anti-tailgating) + Biometric Iris / Palm-Vein MFA.

RING 5 // HALL

White Space & Cage Access: Zone-restricted RFID smart card + PIN credentials limiting engineers strictly to their assigned cage.

RING 6 // RACK

Cabinet-Level Smart Locks: Electronic smart-handle locks on every 42U rack with automated door-open alerts and audit trails.

Cyber Resilience & Governance

Cloud Security — Shared Responsibility & Zero-Trust Defense

Securing workloads across On-Premise and Public Cloud requires mastering the Shared Responsibility Matrix, Zero Trust IAM, encryption, and immutable ransomware recovery.

The Cloud Shared Responsibility Matrix

Security OF vs. IN the Cloud

While cloud providers secure the physical building and hypervisor, Customer Data, IAM Access, Endpoint Security, and Backup Immutability remain the customer's responsibility.

Security Layer	On-Premise DC	IaaS (EC2/VM)	PaaS (RDS/AKS)	SaaS (M365/CRM)
Data, Governance & Backups	Customer	Customer	Customer	Customer
Identity & Access (IAM / MFA)	Customer	Customer	Customer	Shared
Application Code & WAF Rules	Customer	Customer	Shared	Provider
Guest OS, Patching & EDR Agent	Customer	Customer	Provider	Provider
Hypervisor & Virtual Network	Customer	Provider	Provider	Provider
Physical DC, Power & Racks	Customer	Provider	Provider	Provider

01 // Zero Trust

IAM, PAM & ZTNA

Least-privilege RBAC, hardware FIDO2 MFA, Privileged Access Management (PAM) session recording, and isolated Out-of-Band (OOB) management VLANs.

02 // Cryptography

AES-256 & HSM/KMS

At Rest: AES-256 SAN/volume encryption with customer-managed keys (BYOK in FIPS 140-3 HSM). In Transit: TLS 1.3 & IPsec site-to-site VPN.

03 // Perimeter & WAF

NGFW HA & Micro-Seg

Active-Passive FortiGate/Palo Alto NGFW clusters with UTP threat inspection, KEMP WAF load balancing, and East-West VLAN micro-segmentation.

04 // SOC & Endpoint

EDR, CSPM & SIEM

Mandatory Endpoint Detection & Response (EDR) agents on every production VM, Cloud Security Posture Management (CSPM), and 24/7 SIEM log correlation.

05 // Anti-Ransomware

3-2-1 S3 Object Lock

Dedicated Veeam backup server (RAID-6) paired with off-site Wasabi/AWS S3 Object Lock (WORM immutability) so ransomware can never encrypt backups.

06 // Regulatory Law

Global & GCC Frameworks

Full alignment with ISO 27001, SOC 2 Type II, PCI-DSS v4.0, Saudi NCA ECC / SAMA, UAE NESA / ADHICS, and GDPR / PDPL.

Hardware Security Appliances & Continuous Governance SOP

Security Devices Used & How Security is Maintained 24/7/365

A complete inventory of physical, network, and cyber security devices inside the data center—and the operational cadence that keeps them impenetrable.

Fortified 42U Rack Zero-Trust Shield and 24/7 SOC

CONVERGED PHYSICAL + CYBER DEFENSE

Smart Biometric Rack Lock, NGFW HA Cluster, Immutable WORM Vault & 24/7 SOC

Exact Security Devices Installed in the Data Center 12 Defense Appliances

1. PERIMETER & WAF

2× FortiGate FG-201G HA (ASIC NGFW, IPS, AV, SSL Inspection) + 2× KEMP LoadMaster WAF VMs.

2. SWITCH & OOB ISOLATION

Cisco C9300X (802.1X, DHCP Snooping, VLAN ACLs) + Air-Gapped FortiSwitch FS-124F OOB.

3. ENDPOINT & IMMUTABLE DR

EDR/XDR Agents on all VMs + Hardened DL380 Veeam RAID-6 Host + S3 Object Lock WORM.

4. PHYSICAL & HAZARD SENSORS

Biometric/RFID Controller, Smart Rack Locks, AI IR CCTV + NVR, VESDA + FM-200 + EPO.

How Security is Maintained — 5-Stage Operational Cadence

Continuous SecOps SOP

1. REAL-TIME (24/7)

Automated Threat Prevention: FortiGuard IPS/Anti-Malware definitions auto-update hourly; FG-201G HA heartbeat monitors sub-3s failover; SIEM correlates firewall, iLO 7, and biometric door events in real time.

2. DAILY OPS

Backup & Telemetry Verification: Verify 100% completion of nightly Veeam encrypted backups + S3 Object Lock immutability sync; inspect blocked WAN intrusion attempts, UPS load/temp (22°C), and CCTV NVR health.

3. WEEKLY OPS

Access Audit & Restore Test: Audit Active Directory Domain Admin, PAM & iLO 7 login logs; run automated Veeam SureBackup sandbox VM recovery test to guarantee zero backup corruption; review East-West VLAN ACL hits.

4. MONTHLY OPS

Zero-Downtime Rolling Patching: Live-migrate VMs to Host 2 while patching Host 1 (Windows Server, FortiOS, Cisco IOS-XE, HPE Alletra ArcusOS); run Nessus/Qualys CVE vulnerability scans; revoke departed staff RFID badges.

5. QUARTERLY / YR

Live Failover Drills & VAPT Audit: Execute 8-point physical failover drills (unplugging primary firewall/PDU); conduct external Penetration Testing (VAPT); inspect FM-200 gas cylinder pressure & IEEE 1188 UPS battery health.

Preventive & Concurrent Maintenance, OEM Warranties & L1/L2/L3 Support Governance

Data Center Maintenance & 24/7 Support SLA Architecture

How hardware, power, cooling, and software are maintained with zero downtime—backed by multi-year OEM warranties, Annual Maintenance Contracts (AMC), and strict SLA response guarantees.

Data Center Live Hot-Swap Maintenance and 24/7 NOC Support

CONCURRENT LIVE MAINTENANCE & 24/7 NOC

Hot-Swap NVMe/PSU Servicing, Thermal Diagnostics, UPS Bypass & 99.9% SLA Helpdesk

Multi-Year OEM Warranty & Support Coverage

Tier-1 Vendor Backed

HPE 5-YEAR TECH CARE

3× DL380 Gen12 Servers & Alletra MP All-NVMe SAN (24/7 parts, firmware & RMA).

FORTINET 3-YR FORTIGUARD

2× FG-201G HA Firewalls (24/7 FortiCare TAC, UTP threat feeds & hardware RMA).

CISCO & KEMP 3-YR TAC

C9300X / C9200L SmartNet & 3-Year KEMP LoadMaster WAF 24/7 Enterprise Support.

APC UPS, COOLING & VEEAM

APC 20kVA (3-Yr UPS / 2-Yr Battery), InRow PAC & FM-200 AMC, Veeam Production Sub.

Zero-Downtime Maintenance Pillars (Facility, Hardware & Software)

Concurrent Maintenance

1. POWER & UPS BYPASS MAINTENANCE

External Manual Maintenance Bypass Panel switches rack load cleanly to utility/generator so the 20kVA UPS or 192V battery pack can be serviced/replaced with 0ms server downtime.

2. HOT-SWAP COMPUTE, SAN & PDU

Dual Titanium PSUs (PDU-A/B), hot-plug NVMe/SAS drives, and dual-controller Alletra SAN allow failed disks, fans, or power supplies to be swapped live while applications run.

3. COOLING, FLOOR & FM-200 AMC

Quarterly 25.1kW InRow PAC air filter cleaning, condenser coil flushing, refrigerant pressure checks, semi-annual FM-200 cylinder weighing, and underfloor plenum vacuuming.

4. ROLLING CLUSTER PATCHING

Hyper-V Live Migration evacuates VMs to Host 2 while Host 1 receives BIOS/iLO 7/OS patches; StackWise-1T & FG-201G HA enable hitless network firmware upgrades.

SLA Severity Tier	Incident Scope / Trigger	Response & Escalation	Resolution & Support Path
P1 // CRITICAL	Production down, SAN outage, power/cooling failure, active breach	< 15 Mins (24/7 NOC L1/L2) + Instant L3 Architect alert	< 2–4 Hrs On-Site Engineer & OEM Priority RMA; 99.9% SLA
P2 // HIGH	Single HA node/firewall/PDU failure (cluster running degraded)	< 30 Mins Remote OOB (iLO 7 / VPN) diagnostics	< 4–8 Hrs Hot-swap part replacement; zero business stop
P3 // MEDIUM	Non-critical VM issue, backup warning, minor switch port fault	< 2 Hours (L2 System / Network Engineer)	< 24 Hours Remote remediation or next-business-day visit
P4 // PREVENTIVE	Scheduled AMC audit, firmware upgrade, new VLAN/VM request, UAT	< 8 Hours Change Advisory Board (CAB) approval	Planned window + 8-Point Failover Drill & Admin Training

Strategic Architecture Comparison

Private Data Center vs. Public Cloud (AWS, Azure, GCP)

Evaluating On-Premise / Colocation infrastructure against Hyperscale Public Cloud across cost economics, sovereignty, latency, and control.

Private Data Center Connected to Public Cloud

DEDICATED HYBRID INTERCONNECT

Private Data Center Bridging to Hyperscale Cloud via Direct Connect / ExpressRoute

The 3-Year TCO Crossover Rule Financial Reality

Public Cloud excels for spiky or experimental workloads. However, for steady 24/7 enterprise workloads (10–200 VMs, high database IOPS, heavy bandwidth), an On-Premise Mini/Core Data Center typically pays for itself in 18–26 months and saves 40%–55% over 5 years.

CHOOSE PRIVATE / COLO

Core ERP/DBs, strict GCC data residency, predictable 24/7 load, LAN speed.

CHOOSE PUBLIC CLOUD

Global SaaS frontends, seasonal bursts, cloud-native AI APIs, off-site S3 DR.

Comparison Dimension	Private Data Center (On-Prem / Colo)	Public Cloud (AWS / Azure / GCP)
1. Financial Model	CapEx Investment Fixed hardware asset amortized over 5–7 years; minimal monthly recurring cost.	OpEx Subscription Zero upfront CapEx; perpetual monthly billing that scales with usage.
2. Provisioning Speed	4 to 16 Weeks for initial turnkey build; minutes to spin up new VMs once built.	Minutes to deploy new cloud VMs or managed databases via console/Terraform.
3. Data Sovereignty	100% Physical Sovereignty — You control the room, keys, and NVMe drives in-country.	Data sits in shared hyperscaler halls; subject to foreign cloud jurisdiction laws.
4. Scalability	Scale-up by populating free rack units (e.g., 5U buffer) or adding expansion racks/nodes.	Instant Elastic Auto-Scaling across thousands of instances globally.
5. Latency & Egress Fees	<0.5 ms Local 10G/25G LAN; $0 data egress fees for internal file/DB transfers.	Internet/WAN latency (10–60ms); expensive per-GB egress charges when downloading data.
6. Storage & IOPS Cost	All-NVMe 32Gb FC SAN delivers hundreds of thousands of IOPS at zero extra monthly fee.	Provisioned High-IOPS Cloud Block Storage (io2/Ultra Disk) is extremely costly per month.
7. Accountability	Dedicated turnkey integrator (BluePi Solutions) + internal IT control.	Self-service shared responsibility; enterprise support tiers cost 3%–10% extra of cloud bill.

Hyperscaler Comparison & Hybrid Architecture

Private vs. Public Cloud — AWS, Azure, GCP & Hybrid Blueprint

Comparing the Big Three public clouds against Private Cloud stacks, and how enterprises bridge them into a unified Hybrid architecture.

AWS UAE & Bahrain Regions

Amazon Web Services

Strength: 240+ services, mature Middle East regions (ME-Central UAE, ME-South Bahrain).
Core Tech: EC2, S3 (Object Lock), RDS, EKS, AWS Outposts.
Best For: Cloud-native apps, immutable S3 backup targets, global SaaS.

MICROSOFT AZURE UAE, Qatar & KSA Regions

Microsoft Azure

Strength: Seamless licensing & identity sync with Windows Server Datacenter, Active Directory & SQL.
Core Tech: Azure Arc, Azure Stack HCI, Entra ID, ExpressRoute, M365.
Best For: Enterprises running Windows Server Hyper-V clusters & hybrid AD.

GOOGLE CLOUD Dammam (KSA) & Doha

Google Cloud (GCP)

Strength: Industry leader in Kubernetes (GKE), BigQuery analytics, Vertex AI, and Aramco-partnered KSA region.
Core Tech: GKE, BigQuery, Google Distributed Cloud (Anthos).
Best For: AI/ML pipelines, real-time big data warehousing, containers.

PRIVATE CLOUD On-Premise / Colo HA

Private Cloud Suites

Strength: Cloud-style virtualization & HA running on dedicated HPE/Dell + NVMe SAN hardware.
Core Tech: Windows Server Datacenter Hyper-V, VMware VCF, Nutanix, Proxmox/OpenShift.
Best For: Sovereign data control, zero cloud egress fees, sub-ms latency.

The Enterprise Gold Standard: Hybrid Cloud + 3-2-1 Cloud Immutable DR

BluePi Recommended Architecture

ON-PREMISE / COLOCATION PRIMARY DC

Runs core production VMs (Active Directory, ERP, SQL Databases, File Servers, IP-PBX) on HPE ProLiant + Alletra All-NVMe SAN with local RAID-6 Veeam repository for instant minute-level recovery.

ENCRYPTED BRIDGE

↔ TLS 1.3 / IPsec ↔

Direct Connect / ExpressRoute / Veeam Cloud Tier

PUBLIC CLOUD (AWS / AZURE / WASABI S3)

Receives automated off-site S3 Object Lock (WORM Immutable) backups, hosts public customer web portals/CDN, and provides on-demand Cloud Disaster Recovery failover.

Gulf Region (UAE, KSA, Qatar, Oman, Kuwait, Bahrain) Service Catalog

BluePi Solutions in the Gulf — Data Center & Cloud Services

Delivering UK-engineered, turnkey Data Center, Virtualization, Storage, and Disaster Recovery solutions tailored for GCC enterprises and government entities.

GCC Dubai and Riyadh Data Center Network

UK ENGINEERING • GCC EXECUTION

Supporting Saudi Vision 2030, UAE Digital Economy & Qatar National Vision

Why GCC Enterprises Choose BluePi Solutions Single Accountable Partner

Headquartered in Covent Garden, London (UK) under The Integrity Consortium Ltd, BluePi eliminates fragmented vendor finger-pointing by taking single-point turnkey responsibility from raised floor & cooling to SAN, firewall HA, and UAT sign-off.

Riyadh / Jeddah / Dammam (KSA) Dubai / Abu Dhabi (UAE) Doha (Qatar) Muscat / Kuwait / Manama

DC SERVICE 01 // TURNKEY BUILD

On-Premise Mini & Enterprise DC

Complete design & build of 1-rack (42U) to 50-rack server rooms: FS800 antistatic raised floor, entrance ramps, 25kW+ InRow precision cooling (high-ambient 50°C Gulf condensers), and FM-200 / Novec 1230 fire suppression.

DC SERVICE 02 // POWER & UPS

Power Engineering & UPS Autonomy

HPE/Schneider electrical load studies, APC Smart-UPS RT (20kVA–500kVA) online double-conversion deployment, IEEE 1188 battery aging calculations, dual 16A/32A Smart Eco PDUs, and Generator/ATS integration.

DC SERVICE 03 // COMPUTE & SAN

Virtualization & All-NVMe SAN

Turnkey clustering of HPE ProLiant DL380 Gen12 hosts and HPE Alletra Storage MP All-NVMe SAN over dual-fabric 32Gb/64Gb Fibre Channel with Windows Server Datacenter Hyper-V or VMware vSphere HA.

DC SERVICE 04 // COLO & CLOUD

GCC Colocation & Hybrid Cloud

Cage architecture, rack-and-stack, and zero-downtime migration into Equinix Dubai/Abu Dhabi, Khazna, Moro Hub, Center3 (KSA), and Meeza (Qatar) + hybrid ExpressRoute/Direct Connect to AWS/Azure ME regions.

DC SERVICE 05 // BCDR & BACKUP

3-2-1 Veeam Immutable DR

Dedicated physical backup servers (RAID-6 repositories) paired with Veeam Universal Licensing and off-site S3 Object Lock WORM immutability, plus quarterly automated DR failover drills.

DC SERVICE 06 // MANAGED OPS

24/7 NOC, OOB & Annual AMC

Dedicated Out-of-Band (iLO 7, ATEN LCD KVM-over-IP) management networks, 24/7 proactive telemetry monitoring, firmware lifecycle patching, and SLA-backed Annual Maintenance Contracts (AMC).

Comprehensive Cybersecurity, Compliance & Enterprise IT Portfolio

BluePi GCC Portfolio — Security Services & Enterprise Solutions

Complete breakdown of Cybersecurity, GCC Regulatory Compliance, Network Engineering, Unified Telephony, and Custom Software/AI services offered by BluePi Solutions.

A. Cybersecurity & GCC Compliance Services

Zero-Trust Defense

1. Perimeter NGFW HA, WAF & Network Security

Turnkey deployment of FortiGate (e.g., FG-201G) & Palo Alto Active-Passive HA Clusters with Unified Threat Protection (IPS, Anti-Malware, Web Filtering), KEMP LoadMaster WAF/ADC, and dual-ISP SD-WAN demarcation.

2. Core Switching & East-West Micro-Segmentation

High-availability Cisco Catalyst C9300X (StackWise-1T & StackPower) core/server switch stacks, VLAN ACL isolation between DMZ, Production, Guest, and Out-of-Band Management networks.

3. Endpoint Security (EDR/XDR), IAM & PAM Hardening

Enterprise EDR/XDR agent deployment across all production VMs and user PCs, Active Directory Group Policy hardening, MFA enforcement, and Privileged Access Management.

4. GCC Regulatory Compliance & VAPT Audits

Gap analysis, technical remediation, and audit readiness for Saudi NCA ECC, SAMA CSF, Saudi PDPL, UAE NESA (SIA), Abu Dhabi ADHICS, Qatar NIA, ISO 27001, and PCI-DSS v4.0 + Penetration Testing (VAPT).

5. Physical Data Center Security & Access Control

Biometric iris/fingerprint access controllers, electromagnetic locks, AI dome CCTV & NVR retention, and FM-200/EPO relay integration.

B. Enterprise IT, Telephony, Cloud & Software Services

End-to-End Digital Partner

1. Corporate IP Telephony, PABX & Campus PoE+

Turnkey supply & setup of Yeastar P-Series / Cisco IP-PBX appliances, executive IP phones (70–1,000+ extensions), Cisco Catalyst C9200L PoE+ access switching, and local SIP/E1 PRI trunk integration with e& (Etisalat), du, STC, Mobily & Ooredoo.

2. Structured Cabling & Smart Office Fit-Out

End-to-end OM4/OS2 Fiber LIU splicing, Cat6A shielded horizontal cabling, patch panel dressing, Fluke certification testing, and Enterprise Wi-Fi 6E/7 (Cisco Meraki / FortiAP) heat-mapped deployment.

3. Microsoft Licensing, Modern Workplace & ERP Infra

Windows Server Datacenter core licensing, User/RDS CAL governance, Active Directory & DNS architecture, Microsoft 365 / Entra ID hybrid identity, and SQL Server / SAP / Odoo high-availability hosting.

4. Custom Web, Mobile Product & AI Automation

Bespoke enterprise web applications, customer portals, private on-prem LLM/AI assistants (running securely on client GPU/server infrastructure), and automated DevOps CI/CD pipelines.

5. IT Consultancy, Gap Analysis & vCTO Advisory

Independent infrastructure audits, Data Center Gap Analysis, BOM reconciliation, and commercial/technical vendor governance for C-suite boards.

Commercial Growth & Market Penetration Blueprint

How BluePi Solutions Will Market & Win in the Gulf (GCC Strategy)

A 4-pillar Go-To-Market (GTM) and demand-generation playbook to capture high-value Data Center, Security, and Turnkey IT contracts across UAE, KSA, Qatar & the wider GCC.

PILLAR 01 // POSITIONING

UK Engineering Pedigree + "Turnkey in 16 Weeks"

Brand Hook: Position BluePi Solutions (UK) Ltd as a British-engineered turnkey specialist—avoiding Big-4 consulting markups while beating local box-droppers on engineering rigor.
Hero Productized Offer: Market the "Turnkey 42U High-Availability Corporate Mini Data Center" (fixed scope, 16-week delivery, 99.9% SLA) to GCC companies with 50–500 staff.
Sovereignty Angle: Target firms forced by Saudi PDPL, SAMA, and UAE data laws to keep data in-country rather than on foreign public clouds.

Target: KSA, UAE, Qatar, Oman, Kuwait

PILLAR 02 // CHANNEL & JV

MEP Fit-Out, Distributor & Colo Alliances

Commercial Interior & MEP Partners: Partner with Dubai/Riyadh office fit-out and civil contractors—every time a company builds a new HQ floor, BluePi is brought in as the specialist Server Room / Mini-DC subcontractor.
Vendor & Distributor Co-Selling: Register deal protection with HPE, Schneider APC, Cisco, Fortinet & Veeam via Gulf distributors (Redington Gulf, Mindware, Logicom) to receive vendor-referred leads.
Local Presence / JV: Leverage UAE Free Zone (DMCC/DIFC) & KSA MISA/local partner registration for government & semi-govt tenders (Etimad / Tejari).

B2B Partner Referral Multiplier

PILLAR 03 // DEMAND GEN & ABM

ABM, Technical Lead Magnets & SEO

LinkedIn Account-Based Marketing (ABM): Direct executive outreach to CTOs, IT Heads, CISOs, and COOs in Banking, Healthcare, EPC, Logistics, and Real Estate across Riyadh & Dubai.
High-Value Free Diagnostics (Lead Magnets): Offer a free "Data Center Gap & UPS Battery Aging Audit" or "Cloud Repatriation & 3-2-1 Ransomware Readiness Assessment" to get inside client server rooms.
Proof-Driven Content & SEO: Publish real rack elevation blueprints, thermal calculations, and bilingual (English/Arabic) landing pages targeting high-intent GCC search terms.

Audit-First Consultative Selling

PILLAR 04 // EVENTS & EXPANSION

Flagship GCC Events & Land-and-Expand

Flagship Tech Summits: Host private C-level roundtables & bilateral meetings during GITEX Global (Dubai), LEAP (Riyadh), Black Hat MEA (Riyadh), and COMEX (Muscat).
Land-and-Expand Funnel:
1) Land: Security/Gap Audit or FortiGate/Veeam upgrade ($15k–$40k)
2) Expand: Turnkey 42U Mini Data Center or Colo Migration ($150k–$500k+)
3) Retain: 3-Year Managed NOC, SOC & Preventative AMC recurring revenue.

Recurring AMC + Turnkey Margin

BluePi Solutions (UK) Ltd | Subsidiary of The Integrity Consortium Ltd

HQ: 71-75 Shelton Street, Covent Garden, London WC2H 9JQ, UK • Web: www.bluepisolutions.co.uk

Akhtar (Delivery Leadership): akhtar@bluepisolutions.co.uk | +44 (0) 7799861693

Md Amin Abdullah Khan (Tech & Commercial Director): amin@bluepisolutions.co.uk | +44 (0) 7456160450