Adijaya Inc ▮

Panduan Lengkap Master Plan Data Center

Pendahuluan

Data center merupakan jantung infrastruktur digital perusahaan modern. Membangun master plan yang komprehensif adalah kunci kesuksesan implementasi data center, baik untuk skala kecil maupun besar. Artikel ini akan membahas secara detail cara membuat master plan data center dengan standar Tier 2 hingga Tier 4 menurut klasifikasi Uptime Institute.

Memahami Klasifikasi Tier Data Center

Tier 2 - Redundant Capacity Components

  • Availability: 99.741% (22 jam downtime/tahun)
  • Memiliki komponen redundan (N+1)
  • Single path untuk distribusi daya dan pendinginan
  • Maintenance memerlukan shutdown parsial

Tier 3 - Concurrently Maintainable

  • Availability: 99.982% (1.6 jam downtime/tahun)
  • Multiple path untuk daya dan pendinginan
  • Maintenance tanpa shutdown
  • Dual-powered equipment dengan satu path aktif

Tier 4 - Fault Tolerant

  • Availability: 99.995% (26 menit downtime/tahun)
  • Fully redundant systems (2N atau 2N+1)
  • Fault tolerant terhadap single failure
  • Multiple active paths

Fase 1: Analisis Kebutuhan dan Perencanaan Awal

1.1 Analisis Kapasitas

Untuk Skala Kecil (Tier 2):

  • Kapasitas IT load: 50-200 kW
  • Ruang server: 50-150 m²
  • Jumlah rack: 10-30 unit
  • Periode pertumbuhan: 3-5 tahun

Untuk Skala Besar (Tier 3-4):

  • Kapasitas IT load: 500 kW - beberapa MW
  • Ruang server: 500-5000+ m²
  • Jumlah rack: 100-1000+ unit
  • Periode pertumbuhan: 5-10 tahun

1.2 Studi Kelayakan

  • Analisis Lokasi: Akses jaringan fiber, risiko bencana alam, ketersediaan utilitas
  • Analisis Biaya: CAPEX dan OPEX untuk 10 tahun
  • Analisis Risiko: Business continuity dan disaster recovery
  • Compliance: Regulasi pemerintah, ISO 27001, SNI, dll

1.3 Penentuan Target Tier

Pertimbangan memilih tier:

  • Tier 2: Cocok untuk perusahaan dengan budget terbatas, aplikasi non-critical
  • Tier 3: Ideal untuk perusahaan menengah-besar, aplikasi bisnis penting
  • Tier 4: Untuk institusi finansial, healthcare, aplikasi mission-critical

Fase 2: Desain Arsitektur dan Infrastruktur

2.1 Tata Letak (Layout) Data Center

Zona Utama:

  1. Entry/Security Area: Lobby, security control, visitor management
  2. Office Area: NOC (Network Operations Center), SOC, ruang meeting
  3. Support Area: Ruang utilitas, genset, baterai, cooling plant
  4. White Space: Area server dan storage (raised floor)
  5. MDF/IDF Room: Main dan Intermediate Distribution Frame

Desain White Space:

  • Hot Aisle - Cold Aisle Configuration: Standar untuk efisiensi pendinginan
  • Density Planning: 3-5 kW/rack (Tier 2), 8-15 kW/rack (Tier 3-4)
  • Containment: Cold aisle containment atau hot aisle containment untuk high-density

2.2 Sistem Daya Listrik

Tier 2 Configuration:

  • Single utility feed dengan backup genset
  • UPS N+1 (misal: 4 unit UPS untuk kapasitas 3)
  • Single PDU per rack
  • Distribusi: Utility → ATS → UPS → Distribution Panel → PDU → IT Equipment

Tier 3 Configuration:

  • Dual utility feeds (dari 2 gardu berbeda)
  • Multiple UPS systems dengan static transfer switch
  • Dual PDU per rack (A+B feed)
  • Automatic transfer capability

Tier 4 Configuration:

  • Multiple utility feeds dengan routing berbeda
  • 2N UPS configuration (fully redundant)
  • Dual-corded equipment mandatory
  • Compartmentalized power distribution
  • Continuous maintenance tanpa impact

Komponen Kunci:

  • Generator: Capacity N+1 (Tier 2-3), 2N (Tier 4), fuel storage 24-72 jam
  • UPS: Online double conversion, runtime 10-15 menit
  • Battery: VRLA atau Lithium-ion, 10-15 menit backup
  • PDU: Intelligent/monitored PDU dengan remote management

2.3 Sistem Pendinginan (Cooling)

Perhitungan Cooling Load:

  • Total heat load = IT load + lighting + UPS losses + people
  • PUE target: 1.6-2.0 (Tier 2), 1.4-1.6 (Tier 3-4)

Tier 2 Cooling:

  • CRAC/CRAH units N+1
  • Chiller system dengan single path
  • Room-based cooling
  • Cooling capacity: 1.2-1.5x IT load

Tier 3-4 Cooling:

  • Multiple CRAC/CRAH dengan dual distribution
  • Redundant chiller systems (N+1 atau 2N)
  • Row-based atau rack-based cooling untuk high-density
  • Free cooling integration (economizer)
  • Liquid cooling untuk density >20kW/rack

Best Practices:

  • Target temperature: 18-27°C (ASHRAE)
  • Humidity: 40-60% RH
  • Airflow management: blanking panels, brush strips
  • Monitoring: Thermal mapping, sensor deployment

2.4 Infrastruktur Jaringan

Physical Layer:

  • Multiple fiber entry points (diverse path)
  • Meet-me room untuk carrier interconnection
  • Structured cabling: Cat6A/Cat7 untuk copper, OM3/OM4 untuk fiber

Network Architecture:

  • Core-Distribution-Access layer
  • Redundant core switches
  • ToR (Top of Rack) switches untuk setiap rack
  • Out-of-band management network

Bandwidth Planning:

  • Internal: 10GbE minimum, 25/40/100GbE untuk backbone
  • External: Multiple ISP dengan BGP routing (Tier 3-4)
  • Interconnection: Direct peering atau IXP

2.5 Sistem Keamanan

Physical Security (Tier 2):

  • CCTV 24/7 dengan retention 30-90 hari
  • Access control system (card reader)
  • Security guard
  • Mantrap di entry point

Enhanced Security (Tier 3-4):

  • Biometric access control
  • Double mantrap
  • 24/7 NOC monitoring
  • Intrusion detection
  • Vehicle barrier system

Logical Security:

  • Network segmentation (VLAN)
  • Firewall dan IPS/IDS
  • SIEM implementation
  • Regular security audit

Fase 3: Sistem Pendukung

3.1 Fire Suppression System

Tier 2:

  • Pre-action sprinkler system
  • Smoke detection (VESDA)
  • Portable fire extinguisher

Tier 3-4:

  • Clean agent system (FM-200, Novec 1230)
  • Very Early Smoke Detection Apparatus (VESDA)
  • Zoned suppression
  • Integration dengan BMS

3.2 Building Management System (BMS)

Monitoring Parameters:

  • Temperature dan humidity
  • Power consumption (PUE tracking)
  • Water leakage detection
  • Access control logs
  • CCTV integration

DCIM (Data Center Infrastructure Management):

  • Real-time capacity management
  • Asset tracking
  • Change management
  • Workflow automation
  • Predictive analytics (Tier 4)

3.3 Grounding dan Lightning Protection

  • Grounding resistance: <1 Ohm
  • Mesh grounding system
  • Lightning rod dan surge protection
  • Equipotential bonding

Fase 4: Implementasi dan Konstruksi

4.1 Timeline Pembangunan

Skala Kecil (Tier 2): 6-12 bulan

  • Desain detail: 2-3 bulan
  • Procurement: 2-3 bulan
  • Konstruksi: 3-5 bulan
  • Testing & commissioning: 1 bulan

Skala Besar (Tier 3-4): 12-24 bulan

  • Desain detail: 3-6 bulan
  • Procurement: 4-6 bulan
  • Konstruksi: 6-12 bulan
  • Testing & commissioning: 2-3 bulan

4.2 Procurement Strategy

Vendor Selection:

  • Kriteria: Track record, SLA, support capability
  • Multiple vendor untuk redundancy
  • Long-term partnership consideration

Equipment Categories:

  • Critical equipment: UPS, cooling, generator
  • Network equipment: Switches, routers, firewalls
  • Monitoring: BMS, DCIM, CCTV
  • Infrastructure: Racks, PDU, cabling

4.3 Commissioning Process

Testing Stages:

  1. Factory Acceptance Test (FAT): Testing di pabrik vendor
  2. Site Acceptance Test (SAT): Testing setelah instalasi
  3. Integrated Systems Test (IST): Testing integrasi semua sistem
  4. Full Load Test: Testing dengan beban penuh
  5. Disaster Simulation: Testing failover dan recovery

Tier Certification:

  • Design certification dari Uptime Institute (optional)
  • Constructed facility certification
  • Operational sustainability certification (recurring)

Fase 5: Operasional dan Maintenance

5.1 Staffing Requirements

Tier 2:

  • Data center manager: 1
  • Network engineer: 2-3
  • Facility engineer: 2-3
  • Security: 3-6 (shift)

Tier 3-4:

  • Data center manager: 1-2
  • NOC team: 6-12 (24/7 coverage)
  • Facility team: 6-12
  • Security team: 8-16
  • Management staff: 2-4

5.2 Preventive Maintenance

Monthly:

  • Visual inspection semua equipment
  • CRAC filter cleaning/replacement
  • Battery voltage check
  • Generator exercise run (30 menit)

Quarterly:

  • UPS battery test (load test)
  • Cooling system inspection
  • Fire suppression inspection
  • Access control audit

Annual:

  • Generator load bank test
  • UPS full load test
  • Thermographic survey
  • Facility infrastructure audit
  • Disaster recovery drill

5.3 Standard Operating Procedures (SOP)

Critical SOPs:

  1. Change management procedure
  2. Incident response procedure
  3. Maintenance procedure (planned)
  4. Emergency procedure
  5. Access control procedure
  6. Backup dan recovery procedure

5.4 Key Performance Indicators (KPI)

Infrastructure KPIs:

  • Uptime percentage
  • PUE (Power Usage Effectiveness)
  • MTTR (Mean Time To Repair)
  • MTBF (Mean Time Between Failures)

Operational KPIs:

  • Capacity utilization
  • Energy efficiency
  • Change success rate
  • Incident response time

Fase 6: Budget dan Financial Planning

6.1 CAPEX Breakdown (Estimasi)

Tier 2 (100 kW):

  • Civil construction: 30-35%
  • MEP (Mechanical, Electrical, Plumbing): 40-45%
  • IT Infrastructure: 15-20%
  • Security & BMS: 5-10%
  • Total: USD 3,000-5,000 per kW

Tier 3 (500 kW):

  • Civil construction: 25-30%
  • MEP: 45-50%
  • IT Infrastructure: 15-20%
  • Security & BMS: 5-10%
  • Total: USD 5,000-8,000 per kW

Tier 4 (1 MW+):

  • Civil construction: 20-25%
  • MEP: 50-55%
  • IT Infrastructure: 15-20%
  • Security & BMS: 5-10%
  • Total: USD 10,000-15,000 per kW

6.2 OPEX Estimation (Annual)

Fixed Costs:

  • Staff salary: 30-40%
  • Maintenance contract: 15-20%
  • Insurance: 2-5%

Variable Costs:

  • Electricity: 40-50%
  • Cooling (if separate): 10-15%
  • Miscellaneous: 5-10%

Rule of thumb: Annual OPEX = 10-15% of CAPEX

Best Practices dan Lessons Learned

Do’s:

  1. Plan for Growth: Oversize infrastructure 30-50% untuk ekspansi
  2. Modular Design: Build in phases sesuai demand
  3. Energy Efficiency: Invest in efficient cooling dan power system
  4. Documentation: Maintain as-built drawings dan documentation
  5. Training: Regular training untuk staff
  6. Monitoring: Implement comprehensive monitoring dari awal
  7. Vendor Relationship: Build strong partnership dengan key vendors

Don’ts:

  1. Under-specify Cooling: Cooling failure adalah penyebab outage terbesar
  2. Single Point of Failure: Eliminasi SPOF di critical path
  3. Poor Cable Management: Menghalangi airflow dan maintenance
  4. Neglect Testing: Comprehensive testing sebelum go-live
  5. Ignore Documentation: Poor documentation = operational nightmare
  6. Skip Commissioning: Proper commissioning mencegah future issues

Studi Kasus

Case 1: Data Center Tier 2 untuk SME (150m²)

  • Kapasitas: 100 kW IT load
  • Investment: USD 500,000
  • Timeline: 8 bulan
  • Uptime achieved: 99.75% (target 99.741%)

Case 2: Enterprise Data Center Tier 3 (1000m²)

  • Kapasitas: 500 kW IT load
  • Investment: USD 4,000,000
  • Timeline: 18 bulan
  • Uptime achieved: 99.98% (target 99.982%)

Case 3: Mission-Critical Tier 4 (2000m²)

  • Kapasitas: 2 MW IT load
  • Investment: USD 30,000,000
  • Timeline: 24 bulan
  • Uptime achieved: 99.997% (target 99.995%)

Tren dan Teknologi Masa Depan

Emerging Technologies:

  1. Liquid Cooling: Direct-to-chip cooling untuk AI/HPC workload
  2. Renewable Energy: Solar, wind integration
  3. AI-Powered Management: Predictive maintenance, automated optimization
  4. Edge Computing: Distributed micro data centers
  5. Green Data Center: Sustainability dan carbon neutral initiatives

Sustainability Focus:

  • PUE target: <1.3 untuk new builds
  • Renewable energy: 50-100% target
  • Water usage effectiveness (WUE)
  • Circular economy: Recycling dan reuse

Kesimpulan

Membangun master plan data center memerlukan pendekatan holistik yang mempertimbangkan aspek teknis, operasional, dan finansial. Pemilihan tier yang tepat harus disesuaikan dengan kebutuhan bisnis, budget, dan risk tolerance organisasi.

Kunci kesuksesan:

  • Perencanaan matang dengan horizon 5-10 tahun
  • Desain modular untuk fleksibilitas
  • Implementasi best practices dari awal
  • Operational excellence dengan SOP yang ketat
  • Continuous improvement melalui monitoring dan optimization

Dengan mengikuti panduan ini, organisasi dapat membangun data center yang reliable, efficient, dan scalable sesuai dengan kebutuhan bisnis mereka.

Referensi:

  • Uptime Institute Tier Standards
  • ASHRAE TC 9.9 Guidelines
  • TIA-942 Data Center Standards
  • ISO 27001 Information Security
  • SNI (Standar Nasional Indonesia) untuk Data Center
October 31, 2025 ∙ ai transformasi data-center project-manager