Servers, files and the systems the work goes through
A migration is judged by how Monday morning goes.
We rehearse the move, agree a date the business can afford, and test the way back before the day.
Most migrations start with one of these.
- The servers are out of warranty and the renewal quote has just arrived.
- Two businesses have merged and you are paying for two of everything.
- You moved to the cloud and the bill went up instead of down.
If one of these is yours, talk to us before you sign anything.
Contact usWhat we take on.
What we report back
We report what has moved and what is still to go.
- We move servers and the systems your people use every day.
- Shared drives keep their permissions when they move.
- Mailboxes can move between tenants as well, after a merger or a sale.
- Where a platform service can replace a server, we say so before the move.
- Some things should not move yet, and we will tell you which.
Six parts of a migration, and what it costs to skip one.
None of these is technically hard, and each one tends to get dropped when a project runs short of time.
Rehearsing it
If it gets skipped
The first run is on the night it has to work.
What you get instead
The move is run once on real data first, and every problem it finds is written down.
The coexistence window
If it gets skipped
Half the business is on each system, and nobody decided which counts.
What you get instead
Both systems run side by side for an agreed period, and one of them is named as the system of record.
The date
If it gets skipped
It lands on your year end because an engineer was free.
What you get instead
Your people choose the date, and they can move it.
The way back
If it gets skipped
There is a rollback plan on paper, and it has never been tried.
What you get instead
The rollback is tested, and the call to use it is yours.
Telling people
If it gets skipped
The service desk spends the two weeks after the move answering the same questions.
What you get instead
Before the day, people know what changes and who to ring.
Switching off
If it gets skipped
Both platforms are still billing a year later.
What you get instead
The old platform is switched off on a set date, once nothing is using it.
Lifting everything and shifting it.
Every server is copied into the cloud exactly as it was bought.
It then runs all day at that size and bills for every hour, which is why the invoice goes up.
We size each machine on the work it does, and anything that gains nothing from moving stays where it is.
Questions we get asked.
It depends on what is moving and how much of it there is, and you get a written plan with the dates in it before the work starts.
Often we can. A system that cannot run in two places at once needs some downtime, and the plan names those systems so you can work around them.
We roll back to the old platform, which is still running at that point, so you land on something that works.
Sometimes it will, and our costing shows that before you decide. A fair comparison includes new hardware, licence renewals and the day lost when a disk fails.
Where this page sits.
Our work sits in three layers, and each one depends on the one beneath it. The page you are reading is marked.
You are hereMigration & Modernisation
Foundations
Start with what is running today.
The Fortnight is a fixed-scope look at your estate, and the migration plan is built on what it finds. You keep the findings whether or not we do the move.
You get a written scope before anybody commits to anything.