Rack and Stack Services in Mumbai

Mumbai is India's financial capital and the country's most active data centre market, accounting for roughly 35 to 40 per cent of India's total installed capacity. From the Equinix IBX campuses in Airoli to the hyperscale Yotta NM1 campus in Panvel, the Mumbai metro hosts over 90 data centre facilities serving the Bombay Stock Exchange, Reserve Bank of India, NSE co-location, the headquarters of India's largest private banks, and a growing AI infrastructure buildout led by hyperscale cloud providers. Reboot Monkey (EDCS OÜ, registered in Estonia) provides third-party rack and stack services across every major Mumbai data centre — including Equinix MB1 and MB2, NTT Netmagic NM1 and NM2, Yotta NM1 Panvel, GPX MU1 BKC in Andheri, Nxtra by Airtel Vikhroli, CtrlS Andheri, STT GDC Turbhe, Sify Mumbai, and Web Werks Navi Mumbai — under a single service agreement. We are not affiliated with any data centre operator. We are not a hosting provider. We are a physical services specialist that works inside the facilities your equipment is already colocated in, or the facility you are moving into. This page explains exactly what rack and stack in Mumbai entails, what our process looks like from dock receipt through to post-installation documentation, and what makes the Mumbai market distinct for hardware deployment — from India's 230V/50Hz power standard and IS 1293 plug types to DPDPA 2023 compliance, BFSI access protocols, and GPU high-density deployments at Yotta NM1.

Rack and stack is the physical process of deploying new server hardware into a data centre rack from the point the equipment arrives at the facility loading dock through to the moment it is powered on, verified, and handed back to the client ready for remote access. The service covers everything in between: receiving the delivery, checking the manifest, unpacking and inspecting for shipping damage, assembling rail kits, mounting equipment in the correct rack unit positions, connecting A and B power feeds, running structured cabling, labelling every cable at both ends, installing blanking panels to manage airflow, and producing the documentation package that the client, the facility, and any compliance audit may subsequently require. Rack and stack is distinct from smart hands. Rack and stack is a project-based engagement: one-time deployment of new hardware, typically completed within a scheduled maintenance window. Smart hands is an ongoing reactive support service: reseating a failed drive, replacing a transceiver, rebooting a locked server, or carrying out a visual inspection at an engineer's request. Many clients use Reboot Monkey for both, but they are separate service lines with different scopes and SLAs. In Mumbai's BFSI-heavy market, rack and stack frequently involves compliance-sensitive infrastructure — payment switches, core banking servers, NSE co-location trading nodes — where documentation standards are as important as the physical work itself.

Reboot Monkey's engineers cover the full Mumbai metro including both central Mumbai and the Navi Mumbai data centre cluster, which contains the majority of the city's major facilities. **Airoli and Mahape (Navi Mumbai)** Equinix MB1 in Airoli is the highest-priority colocation facility in Mumbai for carrier-neutral deployments. It hosts 280 networks and connects to DE-CIX Mumbai, AMS-IX Mumbai, and NIXI. Rack and stack at MB1 follows Equinix IBX structured cabling guidelines: cables must be labelled at both ends per Equinix's port labelling schema, all unterminated loose cables are prohibited in live customer zones, and fibre patch cords must use LC-LC or SC-LC duplex connectors as specified. All work orders must be submitted via the Equinix IBX Portal at least 48 hours before engineer dispatch. Equinix MB2 in Navi Mumbai carries the same compliance and cabling requirements as MB1. NTT Netmagic NM1 at Mahape is one of Mumbai's oldest and largest data centre facilities, with a strong BFSI tenant base. Rack and stack here often involves high-density blade chassis — HP BladeSystem, Dell PowerEdge M-series, Cisco UCS — alongside legacy infrastructure rationalisation. NTT Netmagic NM2 in Turbhe serves tenants with multi-site footprints across both NM facilities. Managing cabling consistency and documentation standards across both Netmagic sites is a specific value driver for clients running expansion projects. **Turbhe (Navi Mumbai)** STT GDC Mumbai 1 in Turbhe serves enterprise and government tenants in an ISO 27001-certified environment. Rack and stack at STT GDC requires chain-of-custody documentation for all hardware handled, including serial number logging before and after installation. Government tenants may require advance access approvals. Web Werks Mumbai in Turbhe serves SME and mid-market tenants, many of whom lack on-site technical teams — rack and stack projects here are often managed end-to-end from loading dock to power-on verification. **Panvel (Navi Mumbai far)** Yotta NM1 in Panvel is Mumbai's primary high-density and hyperscale campus, founded by Hiranandani Group in 2019 and certified Tier IV by the Uptime Institute. It is the defining site for GPU and AI/ML rack and stack in Mumbai: power densities of 10 to 15 kW per rack are standard, GPU server deployments require high-amperage IEC C19/C20 circuits at 32A, and structured cabling involves MPO-12 and MPO-24 trunk cables for 400G fabric backbones. The Panvel location is logistically efficient for imported hardware cleared through Nhava Sheva port, approximately 15 km away. **Vikhroli and BKC (Mumbai city)** Nxtra by Airtel at Vikhroli sits within Mumbai city limits and serves Airtel-connected enterprise tenants integrating on-net and third-party carrier infrastructure. GPX MU1 BKC in Andheri is a carrier-neutral facility with NIXI Mumbai on-net, and rack and stack there predominantly involves network and peering equipment: Juniper, Cisco, and Arista router chassis where cable labelling must match cross-connect work orders exactly for NIXI and carrier verification. CtrlS Andheri targets BFSI and government tenants and requires work orders submitted through CtrlS's formal change management process before any engineer attends site. Sify Technologies operates out of Rabale in Navi Mumbai, primarily serving legacy enterprise deployments where hardware refresh cycles are common. Note: The Atal Setu Trans-Harbour Link (Mumbai Trans Harbour Link), which opened in January 2024, reduced transit between Mumbai city and the Navi Mumbai data centre cluster to 25 to 35 minutes under normal conditions. This significantly improves scheduling reliability for multi-day rack and stack projects spanning facilities in both areas.

Every rack and stack engagement in Mumbai follows the same structured workflow. Deviations from this sequence require explicit client authorisation. **Step 1: Pre-Deployment Coordination** Before any engineer attends site, Reboot Monkey collects the hardware manifest, rack elevation diagrams, cabling specifications, and any facility-specific work order requirements. For Equinix MB1/MB2, this means submitting the IBX Portal work order 48 hours in advance. For CtrlS, the formal change request must be approved before scheduling. For any facility, engineer names are added to the authorised visitor list with government-issued ID references provided. Customs clearance documentation for imported hardware is verified at this stage. **Step 2: Dock Receipt and Manifest Verification** On arrival at the data centre loading dock, the engineer receives the hardware delivery, signs the freight receipt, and cross-references every carton against the client's purchase order manifest. Serial numbers are photographed and logged before any packaging is removed. Discrepancies — missing units, damage indicators, count mismatches — are reported to the client before proceeding. **Step 3: Unpack and Damage Inspection** Equipment is unpacked in the facility staging area or within the client cage, depending on facility rules. Each server, blade chassis, switch, or storage unit is inspected visually for shipping damage. The front bezel, rail kit components, power cables, and transceiver cage are checked. Any damage identified before installation is photographed and escalated immediately. **Step 4: Rail Kit Assembly and Rack Unit Positioning** Rail kits are assembled and fitted to the rack posts according to the client's rack elevation diagram. Reboot Monkey engineers verify rail kit compatibility against the rack post type — square-hole, round-hole, or threaded — before mounting. Tool-less rail kits (Dell ReadyRails, HP Quick-Release) are standard for 1U to 4U servers. Static rails are used for heavy chassis such as Cisco UCS blade enclosures and NetApp storage arrays, with floor-load capacity verified in advance. Equipment is mounted in the specified rack units and secured. **Step 5: Power Connection and A/B Feed Verification** Power circuits are connected using IEC C13/C14 connectors for standard load equipment and IEC C19/C20 for high-current devices. India's grid operates at 230V AC and 50Hz per IS 1293 (Bureau of Indian Standards), and all Mumbai data centre PDUs conform to this standard. NEMA connectors are not used in Indian data centres. Before powering any equipment, the engineer performs the power verification sequence: reads PDU circuit voltage at 230V AC, confirms amperage headroom against the equipment's rated draw, verifies dual-feed path availability for A+B redundancy where required, and obtains written confirmation from DC operations or a client authorisation before executing power-on. **Step 6: Structured Cabling** All copper patch cabling uses Cat6a for new deployments, terminated to TIA-568B wiring sequence on RJ45 connectors. Fibre patch cords use LC duplex connectors as standard; SC duplex connectors are used where the legacy infrastructure at NM1 or GPX specifies them. For high-density deployments at Yotta NM1 and Equinix MB1, MPO-12 and MPO-24 trunk cables are pulled for 40G/100G/400G fabric backbones, with polarity verified to Method B (key-up to key-up) before insertion. Top-of-rack switching fabrics are connected using 25G SFP28 DAC cables (1m, 2m, 3m), with compatibility confirmed against the switch vendor's transceiver matrix before installation. Active Optical Cable (AOC) is used for runs exceeding DAC reach limits. Power cables and data cables are always routed on separate cable managers — power on one vertical cable manager, data on the opposite side — to maintain separation. **Step 7: Cable Labelling** Every cable — copper patch, fibre patch, DAC, and AOC — is labelled at both ends before being seated. Labels follow the format: [SOURCE-RACK]-[SOURCE-PORT] to [DEST-RACK]-[DEST-PORT]. Equinix's IBX port labelling schema takes precedence inside Equinix cages. Brady or equivalent heat-shrink or flag labels rated for data centre environments are used. Self-adhesive labels are not applied in high-humidity or HVAC airflow paths. **Step 8: Blanking Panels and Airflow Management** Every unused rack unit after equipment installation receives a 1U blanking panel. This is not cosmetic: blanking panels prevent hot-air recirculation in Mumbai data centres' standard hot-aisle/cold-aisle containment configurations. Vertical cable managers and horizontal cable managers are dressed and secured. **Step 9: Rack Grounding and IS 3043 Verification** All racks must be bonded to the facility earth ground per IS 3043, India's Code of Practice for Earthing. Mumbai data centres — including Equinix MB1 and NTT Netmagic NM1 — provide rack-mounted earth bonding points. Ground continuity is tested with a continuity tester after rack installation and before equipment power-on. Failure to verify grounding before power-on creates electrostatic discharge risk on newly installed hardware and is a safety non-compliance in BIS-certified facilities. **Step 10: Power-On Verification and Remote Access Confirmation** Equipment is powered on sequentially. The engineer confirms POST completion, BIOS/IPMI/iDRAC initial accessibility, and — where client configuration permits — network reachability. Any POST errors, power fault indicators, or failed network pings are photographed and reported before the engineer leaves site. **Step 11: Post-Installation Documentation** Within 48 hours of project completion, Reboot Monkey delivers the documentation package: rack elevation diagrams (before and after), cable schedules with both-end labels, power circuit readings per outlet, hardware asset register with serial numbers and rack unit positions, photographic evidence of each installation step, and the signed work order completion report. This package satisfies DPDPA 2023 chain-of-custody requirements, CERT-In 2022 physical change audit log requirements, and the facility compliance requirements of Equinix, NTT Netmagic, STT GDC, and CtrlS.

The 2024 to 2026 AI infrastructure buildout has driven significant rack and stack demand in Mumbai, particularly at Yotta NM1 in Panvel and Equinix MB1 in Airoli, where hyperscale and enterprise clients are deploying NVIDIA DGX H100, HGX A100/H100, and AMD MI300X server systems. High-density GPU rack and stack differs from standard server deployment in several ways: **Power requirements.** A single NVIDIA DGX H100 draws up to 10.2 kW. A rack of 8 DGX nodes approaches 80 kW, which exceeds the capacity of standard rack PDUs. GPU deployments at Yotta NM1 operate at 10 to 15 kW per rack as standard, with some hyperscale configurations running higher. This requires IEC C19/C20 connectors rated at 32A, dedicated high-amperage PDU circuits, and power load verification against the facility's provisioned capacity before any equipment is mounted. Engineers must confirm PDU circuit headroom per circuit — not just per rack — before powering GPU nodes. **Physical handling.** A DGX H100 weighs approximately 60 to 80 kg in its shipping configuration. Moving GPU server systems within a data centre requires appropriate trolleys, lift assistance where aisles permit, and at least two engineers for heavy chassis installation. Yotta NM1's large loading dock and wide cage aisles accommodate this, but advance coordination with facility staff is required for any delivery exceeding 100 kg per unit. **High-density fibre management.** GPU clusters generate the highest port density of any Mumbai deployment type. Spine-leaf networks feeding DGX racks use 400G QSFP-DD DAC and AOC cabling, and the fibre backbone from top-of-rack switches to aggregation layers uses MPO-24 trunk cables. Every port in a GPU cluster must be verified against the fabric topology before the cluster powers on. Misrouted cables in a DGX cluster can invalidate NVLink performance claims and are difficult to diagnose post-installation without correct before-installation documentation. **Thermal verification.** Rear-door heat exchanger (RDHx) checks are performed for deployments where liquid-assisted or direct liquid cooling (DLC) is specified. Liquid cooling preparation at Yotta NM1 includes verifying coolant supply line connections, pressure, and leak indicators before server power-on. For NVIDIA DGX and HGX deployments in Mumbai, Reboot Monkey coordinates directly with the facility facilities team, the client's hardware vendor, and — where applicable — the cooling infrastructure contractor, to ensure the installation window covers power, cooling, and cabling in the correct sequence.

No competitor in the Mumbai rack and stack market specifies what documentation clients actually receive after installation. For enterprise buyers, this matters. For BFSI clients handling payment infrastructure, it is not optional. Reboot Monkey's standard post-installation documentation package for every Mumbai rack and stack engagement includes: - Rack elevation diagrams: before-state (or empty rack state) and after-state, showing every occupied and unoccupied unit position, equipment make/model, and rack unit reference. - Cable schedule: every cable listed with source rack, source port, destination rack, destination port, cable type (Cat6a, LC duplex fibre, MPO-12, 25G DAC), and cable label reference. - Power circuit readings: voltage and amperage measured at every powered outlet, recorded at the time of installation. Circuit ID, PDU ID, and feed (A or B) logged per outlet. - Hardware asset register: every installed device listed with manufacturer, model, serial number, rack unit position, and power outlet assignment. Formatted for direct import into CMDB or asset management tools. - Photographic evidence: installation-sequence photographs at dock receipt, post-mounting, post-cabling, and post-power-on stages. Timestamped and geotagged where facility rules permit. - Signed work order completion report: engineer name, date/time in and out of facility, facility access record reference, and client acceptance signature (remote acceptance via email acceptable). This package is delivered within 48 hours of project completion and is retained in Reboot Monkey's records for the period required by the client's compliance framework. For CERT-In 2022 compliance, tenants must maintain logs of all changes to ICT infrastructure. The Reboot Monkey documentation package provides the complete physical-layer change audit record. For DPDPA 2023, clients whose deployed equipment processes Indian personal data benefit from chain-of-custody records that satisfy data processor obligations under the Act, including documented access controls and handling records.

**India Power Standard: 230V/50Hz and IS 1293** All Mumbai data centres operate on India's national grid standard: 230V AC at 50Hz. Equipment deployed in Mumbai must be rated for 230V/50Hz operation. Power delivery infrastructure uses IEC C13/C14 connectors for standard 10A to 16A loads and IEC C19/C20 connectors for high-current 16A to 32A loads. NEMA connectors — the North American standard — are not used in Indian data centre PDUs. Hardware imported from North American vendors for deployment in Mumbai requires a power compatibility check before installation; equipment rated for 120V/60Hz only cannot be safely deployed without a transformer. PDUs from APC (Schneider Electric), Vertiv, and Raritan are the most common brands across Mumbai facilities. All are stocked in India at 230V/50Hz specification. **Customs Clearance for Imported Hardware** Hardware imported from the United States, Europe, or Asia Pacific for deployment in Mumbai data centres must clear Indian customs before delivery to the facility. Mumbai handles imports through two primary entry points: Jawaharlal Nehru Port Trust (JNPT, also known as Nhava Sheva) for sea freight and Chhatrapati Shivaji Maharaj International Airport (CSIA) for air freight. Hardware is classified under India's HSN (Harmonised System of Nomenclature) customs tariff code. Integrated Goods and Services Tax (IGST) applies to IT hardware imports at rates varying by category. Delays in customs clearance at JNPT or CSIA directly affect rack and stack scheduling. Reboot Monkey coordinates with the client's customs broker to align hardware delivery dates with facility dock booking windows and engineer scheduling. For the Airoli and Mahape cluster (Equinix MB1, Netmagic NM1), loading docks require 48 to 72 hours advance notice for large deliveries. For Yotta NM1 in Panvel, the facility's large-scale loading dock accommodates palletised deliveries with shorter lead times, and the 15 km proximity to Nhava Sheva port reduces last-mile transit risk. **BFSI Access Protocols and NSE Co-Location** Mumbai's BFSI sector generates the largest single category of rack and stack demand in India. The Reserve Bank of India's data localisation directive (April 2018) requires all payment system data to be stored and processed exclusively within India, driving permanent demand for India-resident engineers to perform physical work on payment infrastructure. Reboot Monkey's Mumbai engineer team satisfies this requirement. The NSE co-location service at Mahape, in the same data centre cluster as Equinix MB1 and Netmagic NM1, serves algorithmic trading firms deploying new trading servers. Rack and stack at NSE co-lo facilities must be executed within pre-approved maintenance windows and must observe NSE change freeze periods around market open and close. The Securities and Exchange Board of India (SEBI) IT framework requires market infrastructure institutions to maintain robust IT infrastructure with documented change procedures. **DPDPA 2023 and CERT-In 2022** India's Digital Personal Data Protection Act (DPDPA), enacted in August 2023 under MeitY, classifies organisations that handle personal data on behalf of a data principal as data processors. A rack and stack provider physically handling equipment that stores or processes personal data — including BFSI servers, healthcare records systems, and consumer application infrastructure — falls within this classification. Key practical obligations for rack and stack engagements include: documented hardware serial number logs, chain-of-custody records for all equipment handled, before and after rack elevation diagrams, and the ability to notify the client within 72 hours if a data exposure risk arises during physical handling. The Data Protection Board of India was still being constituted as of early 2026, but compliance preparation under the enacted Act is underway across the BFSI sector. CERT-In Directions 2022, effective 28 June 2022, require organisations to maintain logs of all changes to ICT systems and to report certain incidents to CERT-In within 6 hours. Physical installation and modification of IT infrastructure generates events that must appear in the tenant's physical change audit log. Reboot Monkey's documentation package satisfies this requirement as a standard deliverable, not an optional add-on. NIST Special Publication 800-88 Rev. 1, while a US federal standard, is referenced by multinational tenants deploying hardware at Mumbai data centres that will subsequently be decommissioned or repurposed. Where hardware decommissioning is part of a rack and stack project (hardware refresh, replace-and-remove), NIST 800-88 media sanitisation procedures are applied to storage devices before the equipment leaves the facility. India's E-Waste (Management) Rules 2022 require removed equipment to be channelled to registered e-waste recyclers. For rack and stack projects involving hardware replacement, Reboot Monkey provides e-waste channelisation documentation as a standard deliverable for all decommissioned equipment.

Every other rack and stack provider in the Mumbai market is a data centre operator. Equinix, NTT Netmagic, Nxtra, CtrlS, Yotta, GPX, Sify, and Web Werks all provide installation services for hardware colocated in their own facilities only. If your equipment spans more than one facility, you need a separate contract, separate point of contact, separate documentation format, and separate SLA for each operator. Reboot Monkey operates across all of them. One contract. One documentation format. One SLA. One engineer team that knows the access procedures, work order requirements, cabling standards, and compliance expectations at every facility in the Mumbai metro. This matters most in three scenarios: First, multi-facility deployments. Enterprise clients with footprints across Equinix MB1, Netmagic NM1, and a third facility in the Turbhe cluster — which is a common BFSI configuration in Mumbai — execute a single rack and stack engagement covering all three sites with consistent cable labelling conventions, the same documentation package format, and a single project completion report. Second, hardware refresh and technology migration. Replacing aged infrastructure at one facility while deploying new equipment at another in the same project scope requires a provider who can coordinate both sides simultaneously, with decommission documentation for the old hardware and installation documentation for the new, under a single chain-of-custody record. Third, AI infrastructure buildout. Deploying GPU clusters at Yotta NM1 while simultaneously expanding network capacity at Equinix MB1 for the cloud on-ramp cross-connects that feed the GPU cluster is a pattern Reboot Monkey executes as one coordinated project. Reboot Monkey (EDCS OÜ) is a global third-party data centre services company operating in more than 250 cities across 190 countries. Mumbai is one city in a global network. For multinational enterprises managing data centre infrastructure across India, Europe, and North America, having the same vendor — with the same documentation standards and the same SLA framework — across all locations simplifies procurement, compliance, and operational management. The 4-hour on-site response SLA applies to emergency rack and stack engagements. Planned deployments are scheduled within 48 hours of confirmed booking.