Skip to content

Rack and Stack Services: Server Installation Worldwide

By Reboot Monkey Team

Reboot Monkey engineers physically receive, mount, cable, and verify your hardware inside any datacenter globally. One contract. 250+ cities. No configuration, just deployment done right.

Rack and Stack Services: Server Installation Worldwide

What Is Rack and Stack?

Rack and stack is the physical process of deploying server hardware inside a datacenter facility. It covers everything from the moment your equipment arrives at the loading dock to the point where it is powered on, verified, and handed back to you with full documentation. The service does not include operating system configuration, firmware changes, or network provisioning. Those tasks fall under smart hands. Rack and stack is purely physical: receive, inspect, mount, cable, label, power on, verify, document. For any organisation placing hardware into a colocation facility, this physical deployment phase is where logistics discipline, vendor knowledge, and documentation accuracy matter most. A poorly executed installation creates problems that persist for the lifetime of the equipment: misrouted cables, overloaded PDU circuits, unlabelled assets, and missing serial records that complicate warranty claims and asset audits for years afterward. Reboot Monkey provides rack and stack as a standalone service and as the first physical phase of a broader datacenter migration or new colocation buildout. Engineers are deployed to the facility, not dispatched remotely. Every step is performed on-site and documented with photographic chain-of-proof.
  • Physical hardware deployment from loading dock to powered-on verification
  • Covers receiving, inspection, rail mounting, cabling, labelling, power-on, and POST check
  • No configuration work included (OS, firmware, VLANs remain customer or smart hands scope)
  • Vendor-neutral: works with any hardware manufacturer and any facility operator

The Full Deployment Lifecycle

A rack and stack engagement follows eight sequential stages, each producing documentation that feeds the final installation report. Pre-arrival planning: Reboot Monkey reviews the customer's bill of materials (BoM), confirms rack unit availability, checks per-PDU power budgets, and schedules the site visit. Engineers typically need 48 to 72 hours lead time to coordinate loading dock access and facility credentials. Hardware receiving and inspection: Engineers meet the delivery at the loading dock. Every item is checked for shipping damage, counted against the BoM, and photographed. Serial numbers are captured against the customer's asset register. Rail kit installation and server mounting: Rail guides are installed to the 4-post or 2-post rack frame per the EIA-310-D rack standard (1.75 inches per RU). Servers are mounted and verified level and flush within AFCOM tolerance. Blanking panels fill empty slots to maintain TIA-942 airflow containment. Power and cooling connection: Power cables are routed through cable trays and connected to assigned PDU outlets. Actual power draw at idle is recorded with a clamp meter. Load is distributed across PDU circuits to avoid oversubscription. Network cabling: Cat6A or Cat6 cables run from each server NIC to the designated patch panel ports per the upstream switch assignment. Each cable is labelled at both ends. Maximum cable length is 100 metres per TIA-568 spec. A connectivity test confirms link activity. Labelling and asset tracking: Each device receives an asset tag recording serial number, rack position, PDU outlet, and network port. The final asset register is compiled for the handoff report. Power-on and POST verification: Devices power on in sequence (switches, UPS, storage, servers) with 30-second intervals to avoid PDU inrush overload. Front-panel POST LEDs are confirmed healthy. IPMI access is tested where applicable. Documentation and handoff: The customer receives the complete installation report within 24 hours, covering receiving photographs, serial records, updated rack elevation diagrams, power draw per device, cable labelling logs, and network port assignments.
  • Pre-arrival planning: BoM verification, power budget check, facility access coordination
  • Receiving and inspection: photographic evidence, serial capture, environmental baseline
  • Rail mounting: EIA-310-D standard, AFCOM tolerance (plus or minus 0.25 inches horizontal)
  • Power connection: clamp meter power draw, PDU load balancing, cooling path confirmation
  • Network cabling: TIA-568 compliant, dual-end labelling, connectivity tested before completion
  • Asset tagging: serial, rack position, PDU outlet, network port recorded per device
  • Power-on sequence: staged by device type, POST verification, IPMI access confirmed
  • Handoff report delivered within 24 hours: photos, serial log, rack elevation update, power and network records

Hardware Reboot Monkey Deploys

Reboot Monkey engineers are trained on all standard rackmount form factors and the vendor-specific rail kits that come with them. 1U and 2U rackmount servers are the core of most colocation deployments. Typical examples include Dell PowerEdge R640 and R750, HP ProLiant DL380 and DL560, Lenovo ThinkSystem SR650, and Supermicro A+ systems. Rail kits vary by vendor: Dell ReadyRails, HP Sliding Rails, Lenovo Toolless Rails. Engineers carry the correct rail hardware for scheduled deployments or work with what arrives in the box. 3U and 4U servers are increasingly common as GPU-dense computing enters the colocation market. NVIDIA DGX systems and custom GPU compute nodes require careful handling of heavier chassis (often 30 to 45 kg) and higher power density (1,200 to 3,000 W per unit). Two-engineer lift procedures apply to all chassis above 25 kg. Blade chassis including Dell PowerEdge M-series and HP BladeSystem c7000 occupy 10U to 14U of rack space and require knowledge of blade-specific power interconnects and switch module management. Storage arrays from Dell EMC, NetApp, Pure Storage, and HPE Nimble are installed for physical placement and cabling only. Reboot Monkey verifies drive count and records serial numbers for warranty tracking but does not perform RAID configuration or LUN provisioning unless upgraded to a smart hands scope.
  • 1U/2U/3U/4U rackmount servers (Dell, HP, Lenovo, Supermicro, custom-built)
  • Blade chassis: Dell M-series, HP c7000 (10U-14U, blade-specific procedures)
  • Storage arrays: Dell EMC, NetApp, Pure Storage, HPE Nimble (physical placement and cabling only)
  • Network switches: Cisco, Arista, Juniper, Mellanox (rack placement and link-light verification)
  • UPS modules (6U-9U): two-engineer lift procedure for units above 25 kg
  • PDU distribution: vertical mount, outlet assignment documented and recorded

Vendor-Neutral and Facility-Independent

Reboot Monkey operates as a third-party service provider. The company does not own the facilities where work is performed. This independence is a structural advantage for customers who colocate across multiple operators. On the hardware side, Reboot Monkey's installation process is identical regardless of manufacturer. Whether a customer deploys a standard Dell PowerEdge using Dell ReadyRails or a custom-built 1U server with generic rail brackets, the workflow is the same: receive, inspect, mount to standard, cable to spec, test, document. On the facility side, Reboot Monkey works inside any colocation operator's space: Equinix, Digital Realty, NTT, Colt, Cologix, and every regional and local operator across 250+ cities. Each facility has its own power standards and cabling requirements. Equinix Frankfurt uses 230V IEC 60309 connectors. A North American facility may use 208V three-phase. Reboot Monkey engineers adapt to the local standard without requiring the customer to manage that translation. This vendor-neutral, facility-independent model means a customer can issue a single purchase order for rack and stack work across Equinix FRA6 in Frankfurt, Digital Realty SIN10 in Singapore, and a regional colo in Warsaw, and receive consistent documentation and execution across all three. One contract, one point of accountability, global scope.
  • Independent from all facility operators: Equinix, Digital Realty, NTT, Colt, and any regional colo
  • No hardware vendor preference: identical process for Dell, HP, Lenovo, Supermicro, or custom equipment
  • Adapts to local power standards: 120V/208V North America, 230V/400V EMEA, 110V/230V APAC
  • Single contract across multiple facilities and cities eliminates multi-vendor coordination overhead

Cable Management Standards

Cable management is not aesthetic work. Poorly managed cables obstruct airflow, cause localised hot spots, create tripping hazards during maintenance, and make future hardware moves significantly more expensive. Reboot Monkey applies TIA-942, AFCOM SOP-001, and IEC 61076 cable management standards across all deployments. Power cables and data cables are separated by a minimum of 6 inches to reduce electromagnetic interference. Cables run through trays or overhead raceways and never across the floor. Each cable is labelled at both ends: power cables carry the circuit breaker assignment at the PDU connection point, and data cables are labelled at the server NIC port and at the patch panel port. Cable bend radius is maintained at a minimum of 1 inch for copper Cat6 and Cat6A per AFCOM SOP-001. Velcro cable wraps are used in preference to plastic zip ties, which can fracture under vibration and cause latent damage. Slack of 10 to 15 percent of cable run length is left coiled above the server or behind the rack for future moves. Photographic evidence of cable runs is taken before the final installation is complete.
  • Power and data cables separated by minimum 6 inches to reduce EMI
  • TIA-942: all cables in trays or raceways, no floor mounting
  • AFCOM SOP-001: bend radius maintained, velcro wraps instead of zip ties
  • Dual-end labelling: server port and patch panel port identified at both ends
  • 10-15% cable slack maintained for future moves (AFCOM best practice)
  • Pre-completion photographic record of all cable runs for audit purposes

Compliance Documentation and Chain of Proof

For organisations operating under compliance frameworks, physical deployment documentation is not optional. Rack and stack creates the foundational asset records that auditors expect when reviewing physical security controls. ISO 27001:2022 Annex A.7 (Physical Controls) requires organisations to demonstrate that physical assets are recorded and equipment placement is documented. Reboot Monkey's workflow produces the serial number records, photographic evidence, and asset placement documentation that satisfy these requirements. For environments subject to PCI DSS 4.0, physical security of cardholder data environment (CDE) hardware is a specific requirement under Requirement 9. The chain-of-proof documentation Reboot Monkey produces, including receiving photographs, serial capture logs, and rack elevation diagrams, supports the physical access and asset tracking controls required under PCI DSS 4.0. SOC 2 Type II audits require a documented audit trail for infrastructure changes. The installation report delivered within 24 hours includes the timestamped photo archive, serial number log, and power and network assignment records that satisfy this requirement. All photographs and serial records are delivered to the customer and retained in Reboot Monkey's NOC records for the duration of the engagement.
  • ISO 27001:2022 A.7: asset records, serial capture, and rack placement documentation produced as standard
  • PCI DSS 4.0 Requirement 9: chain-of-proof supports physical CDE hardware tracking
  • SOC 2 Type II: timestamped audit trail for all installation activities
  • 24-hour report delivery: photos, serial log, rack elevation update, power and network assignments
  • Customer data handled as confidential; records delivered to customer and retained in NOC

Who Uses Rack and Stack Services

Three categories of organisation use rack and stack services regularly. Enterprises expanding colocation footprint across multiple cities face a coordination problem: the internal IT team cannot be on-site in Frankfurt, Singapore, and Dallas simultaneously. Reboot Monkey provides the on-site engineering presence in each city without requiring the customer to maintain a global field team. For enterprises under compliance frameworks, the documentation standard described above provides the audit trail that internal IT teams would produce themselves. Cloud-first organisations repatriating workloads from public cloud to owned hardware are entering colocation for the first time. Reboot Monkey handles the entire receive-to-verified workflow, including coordinating with the facility operator on loading dock access, staging area allocation, and access credentials. Hardware vendors and systems integrators that sell server infrastructure but do not employ field engineers rely on Reboot Monkey to handle installation at the customer's chosen facility. This is common for vendors without a global engineering presence who need consistent deployment execution across multiple markets.
  • Enterprises expanding colocation across multiple cities without a global field team
  • Cloud repatriation projects: organisations deploying owned hardware into colocation for the first time
  • Hardware vendors and systems integrators needing facility-level deployment without an in-house field team
  • Mid-market IT teams without engineers experienced in datacenter physical deployment procedures

Rack and Stack vs. Smart Hands: Where One Ends and the Other Begins

The boundary between rack and stack and smart hands is the point at which the hardware is powered on and verified. Rack and stack ends with the POST check confirming the system has booted. Everything that happens after that on the operating system or network configuration side is smart hands work. Specific examples clarify this boundary. Mounting a server into its rack position and connecting it to the PDU and patch panel is rack and stack. Running an OS installation script on that server is smart hands. Physically connecting a network switch to the patch panel and confirming the link light is active is rack and stack. Logging into the switch CLI to configure VLANs is smart hands. For customers who need both, Reboot Monkey can deliver the full sequence in a single site visit when the scope is planned in advance. For customers who only need the physical deployment, rack and stack can be ordered as a standalone service. The customer's own remote team or a third-party configuration provider takes over at the point Reboot Monkey's documentation is delivered.
  • Rack and stack ends at powered-on POST verification with full documentation delivered
  • Smart hands begins at any OS, firmware, or network configuration task
  • Combined deployments (rack and stack plus smart hands) can be executed in a single site visit
  • Standalone rack and stack leaves the system ready for remote or third-party configuration handoff

Global Coverage: 250+ Cities, Any Facility

Reboot Monkey's rack and stack service is available across 250+ cities in 190 countries. Coverage spans the major colocation hubs: Ashburn, New York, Los Angeles, Dallas, Chicago, and Miami in North America; Frankfurt, London, Amsterdam, Paris, Dublin, Stockholm, and Helsinki in Europe; Tokyo, Singapore, Hong Kong, Sydney, and Mumbai in Asia-Pacific; and Johannesburg, Lagos, Nairobi, and Cape Town in Africa. In each deployed city, the 24/7 NOC maintains oversight of scheduled installations. For priority deployments, Reboot Monkey offers a 4-hour on-site response SLA in deployed cities. This applies to P1 incidents during an active rack and stack engagement where a deployment has stalled due to hardware, power, or facility access issues. Customers are notified of any SLA variance before work is scheduled in non-primary cities. Facilities where Reboot Monkey regularly operates include Equinix, Digital Realty, NTT, Colt, Cologix, and hundreds of regional operators. Engineer familiarity with facility procedures means on-site work starts without delay.
  • 250+ cities across 190 countries with active rack and stack capability
  • FLAP cities (Frankfurt, London, Amsterdam, Paris): regular deployments
  • US hubs: Ashburn, New York, Los Angeles, Dallas, Chicago, Miami
  • APAC hubs: Tokyo, Singapore, Hong Kong, Sydney, Mumbai
  • Africa hubs: Johannesburg, Lagos, Nairobi, Cape Town
  • 4-hour P1 SLA in deployed cities for active installation emergencies
  • Works inside any facility: Equinix, Digital Realty, NTT, Colt, regional operators

Pricing Model

Rack and stack services are priced on a per-unit, per-rack, or project basis depending on scope and geography. Per-unit pricing applies to single-device installations or small deployments where the customer needs one or a few servers mounted and connected. This model suits cloud repatriation pilots or hardware expansion into an existing rack. Per-rack pricing applies to full rack buildouts where an entire cabinet is being populated from scratch. This is the most common model for new colocation deployments and hardware refresh cycles. Project-based pricing applies to multi-rack, multi-city, or multi-facility deployments with coordinated scheduling and a project manager overseeing execution across locations. All pricing is quoted after a scope review covering hardware types, facility locations, access requirements, and documentation deliverables.
  • Per-unit pricing: single-device or incremental installations into existing racks
  • Per-rack pricing: full cabinet buildouts for new colocation deployments
  • Project pricing: multi-rack or multi-city engagements with coordinated scheduling
  • All pricing quoted after scope review covering hardware types, locations, and documentation requirements

Rack and Stack: Common Questions

What exactly is included in a rack and stack service?

Rack and stack covers the full physical deployment lifecycle: hardware receiving and inspection at the facility, unboxing and BoM verification, rail kit installation, server mounting, power cable connection to PDU, network cabling to patch panel, asset tagging and serial capture, power-on and POST verification, and a written installation report delivered within 24 hours. It does not include OS installation, firmware updates, network configuration (VLANs, routing), or any other configuration work. Those tasks fall under smart hands.

How long does a rack and stack installation take?

A standard 1U or 2U server takes approximately 30 to 60 minutes to receive, inspect, mount, cable, and verify. A full rack buildout of 20 to 40 devices typically takes one full working day for a two-engineer team. Blade chassis and storage arrays require additional time due to their complexity. Multi-rack or multi-facility projects are scoped individually based on hardware count, facility access hours, and geographic spread.

Can Reboot Monkey work inside any datacenter facility?

Yes. Reboot Monkey operates as a third-party provider and works inside any facility: Equinix, Digital Realty, NTT, Colt, Cologix, and regional and local operators across 250+ cities. The customer arranges facility access authorisation. Reboot Monkey handles the rest. Engineers adapt to each facility's specific power standards, cabling requirements, and access procedures.

What documentation is delivered after the installation?

The installation report delivered within 24 hours includes: receiving photographs (condition on arrival), serial number log matched to the customer's asset register, updated rack elevation diagram showing new equipment positions, power draw measurements per device, PDU outlet assignments, cable labelling log (server port to patch panel port mapping), network connectivity confirmation, and front-panel POST status. This documentation supports ISO 27001:2022 A.7 asset management, PCI DSS 4.0 Requirement 9 physical hardware tracking, and SOC 2 Type II audit trail requirements.

What is the difference between rack and stack and smart hands?

Rack and stack is physical deployment: the hardware is received, mounted, cabled, powered on, and verified operational. Smart hands begins at the point of configuration: OS installation, firmware updates, VLAN setup, routing changes, or any task that requires logging into the device. If you need both, Reboot Monkey can perform rack and stack followed immediately by smart hands in a single site visit when the scope is agreed in advance.

Does Reboot Monkey handle GPU servers and high-density hardware?

Yes. Reboot Monkey engineers are trained on 3U and 4U GPU-dense systems including NVIDIA DGX servers and custom compute nodes with power draws of 1,200 to 3,000 W per unit. Two-engineer lift procedures apply to chassis above 25 kg. Power connection for high-density hardware is treated as a separate verification step, with clamp meter measurements confirming the actual draw does not exceed the PDU circuit rating before the deployment is signed off.

How is rack and stack priced?

Pricing is structured as per-unit (for single-device or incremental installations), per-rack (for full cabinet buildouts), or project-based (for multi-rack or multi-city deployments). All pricing is quoted after a scope review covering hardware types, facility locations, access requirements, and documentation deliverables. Contact Reboot Monkey to request a quote.

What compliance standards does Reboot Monkey's installation process satisfy?

The rack and stack workflow produces documentation aligned with ISO 27001:2022 Annex A.7 (physical asset management and placement records), PCI DSS 4.0 Requirement 9 (physical security of CDE hardware with chain-of-proof documentation), SOC 2 Type II (timestamped audit trail for infrastructure changes), TIA-942 (structured cabling and cable management), and EIA-310-D (rack dimensioning and equipment mounting standards).

Plan Your Rack and Stack Deployment

Request a Quote