India’s Leading Domain & Web Hosting Provider
Support
Domain & Login
Blog
Data Centre to Cloud Migration: Exit Racks, Contracts and AMCs on Time
Data centre to cloud migration is the programme that empties an owned server room or colocation cage into a public cloud by a contractual exit date, then decommissions the hardware. NTSPL plans the exit backwards from lease, AMC and link notice dates, moves servers in rack-aligned waves to AWS, Azure or Google Cloud, and certifies every disk as wiped.
Delivered on
In scope
What a Data Centre Exit Actually Has to Clear
Most Indian data centre exits are triggered by a date nobody controls: a colo renewal with a 12-18% escalation, a UPS AMC the vendor will not extend, or a landlord reclaiming the floor. The servers are rarely the hard part. The hard parts are the forgotten items in the cage: a licence dongle on a USB hub, a tape library holding seven years of statutory records, and a leased line with a 90-day exit notice. Miss one and the cage cannot be surrendered, so you pay for two estates.
NTSPL runs the exit as a facilities-and-contracts project with a migration inside it. A physical walk-down tags every powered device to a rack unit, owner and contract. The exit calendar works backwards from the surrender date, and waves are sequenced so whole racks go dark together, releasing power, cooling and contracts early. Decommissioning to NIST SP 800-88 is part of scope.
When a Full Exit Beats a Partial Move
Signals you need Data Centre to Cloud Migration
- Colo renewal is 6-12 months out and the quote rises more than 10% per rack.
- Storage arrays or blade chassis lose vendor support within 18 months and the refresh quote exceeds INR 1.5 crore.
- The server room had two or more power or cooling incidents in a year.
- Fewer than 300 servers remain after earlier SaaS moves, so per-rack overhead now exceeds per-VM cloud cost.
When to choose something else
- If over 40% of the estate is plant-floor OT with sub-10 ms loops, keep an edge room and migrate only IT workloads.
- If there is no facility deadline and only a few servers to move, use Server & Workload Migration instead.
- If the exit is under 10 weeks away for 150+ servers, negotiate a short colo extension first; orphans cost more than the extension.
Facility, Contract and Workload Workstreams in an Exit
Walk-down and asset tagging
Every rack unit is photographed and reconciled with switch ports and power strips, so appliances, dongles and modems get an owner and exit action.
Contract and notice register
Colo, cross-connect, leased line, AMC and DR-site contracts are logged with notice periods and tied to the rack group whose move releases them.
Interim hybrid connectivity
Dual VPN or a temporary dedicated link for the 3-9 months of split estate, sized from measured east-west traffic.
Rack-group waves
Waves are grouped by rack and dependency so each one darkens physical racks and lets circuits be billed down.
Decommissioning and disposal
Disks are wiped or shredded with certificates, leased kit returned against serials, and e-waste sent to an authorised recycler.
Rack-to-Region Exit Pipeline
Rack-to-Region Exit Pipeline with Contract Release Gates
Physical and logical inventories merge into one exit register that drives rack-group waves over an interim link into an Indian-region landing zone. Each emptied rack passes a gate that releases power, cross-connects and AMC lines.
Select any component
Rack-to-Region Exit Pipeline with Contract Release Gates: click any of its 8 components to see its role and the design choice NTSPL makes for it, or trace the 7-step primary path.
Exit Tooling on AWS, Azure and Google Cloud
MGN carries servers and DataSync carries file shares over the same link into Mumbai, throttled per wave; Hyderabad holds DR copies.
One Azure Migrate project covers VMware, Hyper-V and physical servers, which suits mixed server rooms. Central India is the usual primary.
Migration Center provides inventory and cost views; Migrate to Virtual Machines runs waves into Mumbai with Delhi as second region.
Exit Decisions That Set the Surrender Date
Full exit versus edge cabinet
Keep a 6-12U edge cabinet only for a named workload needing under 10 ms to plant equipment, survival during WAN loss, or hardware that cannot be virtualised. A retained cabinet still needs UPS, cooling and access control, roughly INR 4-8 lakh a year, so assign an owner and budget before the exit plan is approved.
Interim link sizing
1 Gbps moves roughly 8-9 TB a day at 80% utilisation. Below 40 TB of initial data, dual VPN over two ISPs is enough. Above 100 TB, or where chatty database and app pairs will be split across sites for weeks, take a 6-12 month dedicated link with an exit clause aligned to the surrender date.
Easy-first or rack-first ordering
Easy-first ordering leaves half-empty racks still drawing power. NTSPL runs one or two pilot waves of low-risk servers, then switches to rack-first ordering so each later wave darkens racks. If a critical system shares a rack with test servers, move the test servers early and give the critical one its own planned weekend.
When to serve notice
Serve notice only after the dependent rack group has been dark for 14 days. For a colo with 90-day notice, the last production wave must finish at least 105 days before surrender. Links and AMCs with 30-day notice can be released wave by wave, which typically trims double-running cost by 20-30%.
The 20-to-30-Week Data Centre Exit Sequence
Walk-down, contract register and exit calendar
Physical audit, dependency capture and a calendar built back from the surrender date.
Landing zone and interim link
Accounts, network hub, identity and logging built; link tested at full replication load.
Pilot waves
Two waves of 10-20 low-risk servers prove runbooks and rollback.
Rack-first production waves
Fortnightly waves of 25-60 servers, each ending with rack-dark checks and contract release.
Tapes, sanitisation and surrender
Archives restored to cloud tiers, disks wiped, hardware returned, cage inspected.
What you receive
Exit Progress Measures the CFO Will Ask For
| KPI | Typical target |
|---|---|
| Racks darkened versus plan | Within 1 rack each fortnight |
| Double-running period | Under 4 months below 300 servers |
| Untagged devices found after week 4 | Zero |
| Assets certified at surrender | 100% |
Where Data Centre Exits Slip, and the Countermeasures
| Risk | How NTSPL prevents it |
|---|---|
| An unowned device, such as a licence server or ISP modem, blocks surrender in the final month. | Switch ports and power strips are reconciled in week 2; unowned ports are shut for 48 hours to find the owner. |
| Colo notice served too early forces a rushed last wave. | Notice is gated on rack-dark plus 14 days, with a month-by-month extension rate pre-negotiated. |
| Statutory records on tape become unreadable once the library leaves. | Tapes under Companies Act or GST retention are restored and verified in cloud archive storage before the drive ships. |
Data centre exit to cloud or move to a cheaper colocation provider?
Move colo only if most servers are under three years old, you have staff to run hardware, and load is steady all year; then a colo move is cheaper in year one. If hardware is older than four years, a storage refresh is due, or you run fewer than 300 servers, cloud usually wins within 24-30 months because you skip refresh capex, AMCs and a second DR site. A colo-to-colo move also repeats the physical move risk with no lasting change in how you operate.
Key facts: Data Centre to Cloud Migration
Data Centre to Cloud Migration in Practice
2 Data Centre to Cloud Migration scenarios, from Steel trading and distribution and Private university, plus related platform case studies.
Steel trader surrenders its colo cage 11 days early and drops INR 62 lakh a year
Private university clears its campus server room before a new academic block is built
EPC company vacates its owned data centre in 7 months and halves run cost
Specialty chemicals group links 6 plants and HQ to AWS with dual-path hybrid network
Data Centre to Cloud Migration: Frequently Asked Questions
How long does a data centre exit to cloud take?
For 100-300 servers, plan 20-30 weeks from walk-down to cage surrender. Six weeks go to inventory, contracts, landing zone and link; production waves then run fortnightly. Contract notice periods and change freezes set the length as much as server count does.
What does a data centre exit cost?
Migration services for a 150-server exit typically land between INR 45 lakh and 1.2 crore, depending on database complexity and weekend cutover hours. Add 3-4 months of double running and the interim link, which we report weekly so the CFO sees it shrink as racks go dark.
Does moving to an Indian cloud region satisfy DPDP Act 2023 and CERT-In rules?
Indian regions keep data in-country, but the obligations concern controls, not just location. We set 180-day in-country log retention for CERT-In, configure alerting that supports the 6-hour reporting window, and hand over a map of personal data stores and processors that your DPDP accountability records can reference.
What happens to our tape backups?
Tapes holding records you must keep, such as eight years of accounting books under the Companies Act, are restored and verified into a cloud archive tier before the tape drive leaves the cage. Tapes past retention are degaussed or shredded, with a certificate for each barcode.
Can some servers stay on-site after the exit?
Yes, with a clear reason such as a plant system needing sub-10 ms response or hardware that cannot be virtualised. Retained items go into a small edge cabinet with its own UPS and owner, costed separately so the business sees its real annual cost.
Who handles disposal of old hardware?
NTSPL coordinates it. Disks are wiped to NIST SP 800-88 or shredded, leased equipment is returned against serial lists, and owned hardware goes to an authorised e-waste recycler. You get a certificate pack mapping each serial to its disposal method.