Data centers move. Workloads migrate. I have done it 40+ times. Here is what I have learned.

Pick-and-pack moves are the most under-scoped projects I see. Clients call thinking it is two days of cabling and a weekend cutover. It is rarely that simple. The order of operations matters more than the cabling.

What we plan first: the network path. New site needs the right carrier circuits provisioned, terminated, and tested before any equipment moves. This is the single biggest delay in 80% of moves I have inherited from someone else.

Then equipment. Inventory, label, photograph, ship, rack, recable. Coordinate the power-up sequence against your application dependency map (if you have one; we build one if you do not). Typical small-to-mid move: 1 to 2 days of cutover work, plus 2 to 4 weeks of pre-work.

Most Ohio mid-market companies that move to colocation do it for one of three reasons: their office HVAC cannot keep up with the rack heat, their building does not have generator + UPS in a usable configuration, or they need carrier diversity that their current building cannot deliver.

For Columbus-area moves, we work primarily with Cologix (CMH1, CMH2), Expedient (Columbus, Dublin), Hurricane Electric, and the major regional carrier facilities. Each has trade-offs, carrier mix, cross-connect pricing, power density, remote-hands availability, that matter more than the brand name.

We help you compare facilities against your actual workload profile, negotiate the cabinet contract (which is full of MAC-fee and power-overage clauses people miss), and run the physical move.

Lift-and-shift is moving an existing workload to cloud infrastructure with minimal changes. Fast. Cheap. Wastes most of the cloud benefit. Good for older applications that are stable, low-touch, and not worth re-architecting.

Modernization rebuilds the workload to take advantage of managed services, serverless, autoscaling, the things you actually pay cloud for. Slower. More expensive upfront. Cheaper to operate long-term if the workload justifies it.

The honest read: about 60% of what I see clients move to cloud should have been lift-and-shift. Another 20% should have stayed on-prem. The remaining 20% is where modernization is worth the engineering investment.

Cloud migrations between providers happen more than people expect. M&A activity drives most of them. Cost pressure drives the rest. The technical work is largely tooling (Velostrata, AWS DMS, Azure Migrate, similar). The hard work is the contract negotiation on the way out and the way in, and the data egress costs that nobody quotes accurately upfront. I read those contracts before you sign them.

What I will not sell in data center work

• ×A move plan that does not include network path provisioning at the new site before equipment ships. This is the single most common reason moves slip 2-4 weeks.
• ×A cloud migration sold on cost savings without a documented total cost of ownership across compute, storage, egress, and managed-service consumption over 36 months.
• ×A colocation contract without explicit cross-connect pricing, MAC-fee schedule, and power overage rates spelled out. The defaults will hurt you.