Cloud Consulting/Cloud Migration
Build · Cloud Consulting

Cloud migration, with zero downtime

From on-premise to cloud, without disruption. Audit, roadmap and controlled cutover to a robust, resilient infrastructure, with zero service interruption.

0Service interruption
Lift & refactorDepending on context
Multi-cloudAWS - Azure - GCP - Alibaba

What it covers

  • Audit of the existing estate and migration roadmap
  • Lift & shift or refactoring migration
  • Progressive cutover with no interruption
  • Post-migration optimization and FinOps

How we work

Understand first

We start from your real estate and constraints, not a theoretical model.

Cut over without disruption

Progressive, reversible, tested migration. Production never stops.

Cloud migration: definition

Cloud migration means moving your data, applications and workloads from servers you run on site (on-premise) to the infrastructure of a cloud provider such as AWS, Azure or Google Cloud. You no longer manage the hardware or the lower layers: you use capacity and managed services, billed by usage.

What you gain: elasticity, capacity follows the load instead of being bought for the peak; a readable cost per service and per environment; resilience, several datacenters per region and a failed machine replaced without downtime; proximity to your users, wherever they are. What you hand over: direct control of the lower stack, offset by rigorous architecture and operations.

Which path for each application

A cloud migration is not decided as a block. Each application follows its own path, chosen during the audit according to its criticality, coupling and remaining lifetime.

Rehost (lift & shift). The application moves as is onto cloud virtual machines. The fastest way out of a datacenter, optimising afterwards.

Replatform. Few code changes, but the database, storage or runtime move to managed services. Part of the operations work disappears without a rewrite.

Refactor. The application is reworked to use the cloud: containers, serverless, elasticity. Longer, kept for the applications that carry the business and will last.

Retire or retain. Part of the estate is no longer used and is switched off; another part stays on site for a real reason (regulatory constraint, specific hardware) and is connected to the cloud.

To leave a whole datacenter, see our datacenter-to-cloud migration offer.

Our method, from audit to operations

1. Audit of the existing estate. Inventory of servers, applications, flows and dependencies, criticality of each domain, security and compliance constraints.

2. Priced roadmap. One path per application, migration waves ordered from the least coupled to the most critical, and the target running cost in the cloud.

3. Cloud foundations. Accounts, network, identities, logging and backups set up once, as infrastructure as code, before the first application.

4. Progressive cutover. Data replication, parallel run, verification, then traffic cutover with rollback possible until the last moment.

5. Optimisation and operations. Right-sizing, switching off non-production environments, FinOps, then 24/7 managed services if you want them.

Data and production cutover

One owner. An architect carries the migration end to end: paths, wave order, data plan, cutover criteria. Without one, each team migrates its own way.

The data migration plan. The most delicate step. Depending on the database: two-way synchronisation between site and cloud during the transition, or one-way replication (AWS DMS) and then the old source is cut off. Files, media and logs go through the provider's transfer services, such as AWS DataSync, Storage Gateway or Direct Connect for large volumes.

Big bang or progressive cutover. A big bang moves a whole block during a downtime window: simple to run, but rollback is often impossible. A progressive cutover runs source and target in parallel and moves traffic in steps: no downtime, rollback stays possible. It is our default; the big bang is kept for applications that cannot run two active sources.

One cloud or several. Multi-cloud is justified by a real constraint (regulatory, contractual, a service missing elsewhere), not by principle: it doubles the tooling and skills to maintain.

Cost and funding of a cloud migration

The cost of a migration depends on the number of applications and how tightly coupled they are far more than on the size of the infrastructure. The audit prices it path by path, and compares today's running cost with the target cost in the cloud.

A migration does not cut the bill by itself: a like-for-like rehost often costs more than the original datacenter. The savings come from what follows, once sizing is adjusted and managed services replace machines maintained by hand.

Part of the cost can be funded: the AWS MAP programme covers a share of the assessment and of the move out of a datacenter. Access goes through an AWS partner.

Frequently asked questions

How long does a cloud migration take?

It depends on scope: an audit first, then a domain-by-domain path. Our reference: in late 2017 we migrated our own datacenter and moved 100 clients over with zero interruption.

Will there be downtime?

The target is zero downtime: progressive, reversible, tested cutover. Production never stops.

Lift & shift or refactoring?

Both, depending on context. Lift & shift is fast; refactoring exploits managed services. We price both paths in the audit.

Which applications should a migration start with?

The least coupled and least critical ones: they validate the target, the tooling and the cutover procedure without exposing the core of the business. The most intertwined applications go last, with a team that has already done it.

Does a cloud migration lower costs?

Not by itself. A like-for-like rehost often costs more than the original datacenter. The savings come after the migration: right-sizing, switching off non-production environments, managed services instead of maintained machines.

Is there funding for a cloud migration?

Yes. Cloud providers run programmes that cover part of the cost of assessment and of moving out of a datacenter, such as AWS MAP. Access goes through a partner. We are an AWS Select partner.

AWS, Azure or GCP: which cloud should you choose?

The one that fits your applications, your in-house skills and your data-location constraints, not a choice of principle. We migrate to AWS, Azure, GCP and Alibaba Cloud, and the audit compares the options on your case.

Further reading

On Scale Me Now, our partner consulting firm:

Where we operate

We operate your platforms worldwide, in every cloud region, with no geographic limit.

MigrationGCP→AWSRefactoringWell-Architected
/ They trust us
Premaccess is driving the modernization of our stack to managed AWS services, domain by domain, without ever blocking production. We move forward on the product, they secure the scale-up.
Olivier Thierry
Olivier ThierryCTO · Mobile Club (translated from French)
/ Cloud Migration

Need help with cloud migration ?

Talk to an expert

An expert replies within 24 business hours.