Managed service plans
We run your environment, with 24/7 monitoring, maintenance, and proactive improvement.
Retiring 75 of 100 servers before moving any of them
In Microsoft Azure, consumption equates to cost, so an over-provisioned estate costs more in the cloud than it does on the floor. This one was right-sized first, and the saving is what the right-sizing was worth rather than what the migration was.
Monthly cloud spend removed
$65,000
Off the Microsoft Azure bill after workloads were resized, and sustained after cutover.
- Servers retired before migrating
- 75 of 100
- Databases in the estate
- 1,000+
- To a stable estate
- 6 months
In short
Why the order mattered
A large financial firm responsible for over $35 billion in assets wanted to modernize its SQL Server estate and move it to the cloud. A data tier assessment found data sprawled over 1,000 databases across 100 separate servers, some over-utilized and others essentially asleep. Fortified Data organized the data, resolved duplicated databases, updated legacy SQL Server versions and built a right-sized estate on premises first, decommissioning 75 of the 100 servers. Only then did the migration to Azure begin. Monthly cloud consumption came down by $65,000 and stayed down.
$65,000
Lower monthly Microsoft Azure consumption after workloads were resized and duplicate servers consolidated.
Fortified Data case study, 2025
75 of 100
On-premises servers decommissioned before any workload moved, leaving 25 running.
Fortified Data case study, 2025
6 months
To a fully stabilized environment: every server healthy, data protected, maintenance processes deployed.
Fortified Data case study, 2025
Why an over-provisioned estate is more expensive in the cloud
When a cloud migration is on the table, the first thing worth understanding is the detail of every workload in the environment. Organizations often have mismatched estates, with some servers over-utilized and others essentially asleep. On premises that is a capacity problem. In Microsoft Azure consumption equates to cost, so the same estate carries its waste across and then bills for it every month.
This firm was eager to move, and the initial data tier assessment is what established what was actually there: data sprawled over 1,000 databases in 100 separate servers. The size of that environment decided the first objective, which was a right-sized, optimized, highly available estate delivered before any migration began.
Utilization management in the cloud is not a single dial either. Network pipe, disk space usage, in-out operations and CPUs all attribute to performance and availability, and all of them attribute to the bill.
The engagement, in order
Five steps, and the order is the method
Every step below happened on premises until the last one. That is the whole of the argument: the saving came from what was retired before the migration, not from the migration.
-
Assess before agreeing to anything
A data tier assessment collected trending activity data across the estate and found data sprawled over more than 1,000 databases in 100 separate servers, some over-utilized and others essentially asleep. Capacity decisions were made against measured workload from that point on, rather than against the allocation the environment happened to be carrying.
The baseline every later figure is measured against. -
Consolidate and upgrade on premises
Fortified Data organized all customer data, managed multiple duplicated databases and updated legacy SQL Server software. The team expanded the capabilities of features and applications, increased the performance of the operating system underneath them, and reduced the infrastructure size.
Duplicate databases resolved and legacy versions brought current, before anything moved. -
Decommission what the data said was redundant
With a right-sized architecture in place, 75 servers were decommissioned, leaving 25 on premises. This is the single decision the rest of the outcome rests on, and it happened while the estate was still on the floor, where retiring a server costs nothing per month either way.
75 of 100 servers retired, with no loss of service. -
Within six months
Stabilize, then move
Within six months the environment was fully stabilized: all servers healthy, data protected, and maintenance processes deployed. Only then did the team begin reviewing the detail of the first cloud migration project.
6 months to a healthy, protected, maintained estate. -
Migrate, and stop it drifting back
Workloads were resized, duplicate servers consolidated and redundant SQL databases decommissioned as the estate moved to Azure and on to IaaS operations. DevOps templates, integration, automation and orchestration were implemented alongside, so nothing enters the cloud environment except through the scripted, self-serve portal.
$65,000 a month off Azure consumption, sustained after cutover.
75
What the order was worth
Of 100 servers, retired before anything reached Azure. Hardware costs and operating systems came out of the environment with them, and the monthly cloud bill was set against 25 servers rather than 100 from the first day of the migration.
What changed
- Server sprawl reduced from 100 to 25On premises, and before the migration, which is what set the size of everything billed afterwards.
- Monthly cloud spend reduced by $65,000Measured against the same environment's pre-engagement run rate, and sustained after cutover rather than at the moment of it.
- Hardware and operating systems out of the estateThe move to fully cloud-based IaaS operations removed the hardware cost of the 25 that remained and the operating systems under them.
- Nothing enters the cloud unscriptedDevOps templates, integration, automation and orchestration mean new workloads arrive through the self-serve portal, which is what stops a right-sized estate drifting back.
Why right-sizing came first
Right-sizing the environment to understand workload and availability is key not only to server performance, but it also deeply impacts cost in the Cloud. We have a myriad of tested and trusted tools that help us right-size the environment based on trending data. That was a big contributor to the cost savings.
Three ways to work with us
On-Call DBA
A guaranteed 15-minute emergency response plus monthly on-demand hours that bank when unused.
Database consulting and projects
Project-based engagements for migrations, performance tuning, data modernization, business intelligence, and high availability and disaster recovery.
Migrating an estate you have not measured?
The question worth answering before a migration date is which servers are carrying real workload and which are asleep. On premises that answer saves capacity. In Azure it saves a monthly bill, every month, for as long as the estate runs.
Let us show you what's possible.