Datacentre Exit With a Fixed Deadline: Executed, Not Advised
A logistics and customs brokerage group (340 staff) had nine months to exit its colocation facility — the provider was closing it, with no renewal option. Two racks and 34 physical servers ran the customs declaration platform the entire business depended on: ~12,000 declarations daily, three national customs authority integrations, and hard filing windows where misses carry per-declaration penalties and risk to the broker authorization. Envion ran an embedded migration: leading architecture and cutover planning while the client’s four engineers executed and took ownership. Delivered in month 7 with a 47-minute cutover, zero declarations delayed, zero data loss — and no ongoing Envion involvement, by design.
The challenge
Their colocation contract was expiring in nine months and the provider was exiting the facility, so there was no renewal option. Two racks, 34 physical servers, a mix of virtualized and bare-metal, running the customs declaration platform their entire business depended on.
The platform processed roughly 12,000 customs declarations daily, integrated with three national customs authority systems, and had a hard operational constraint: customs filing windows. Declarations must reach the authority within defined periods, and missing them carries per-declaration penalties and, in repeated cases, risk to the client's customs broker authorization. There was no maintenance window longer than about two hours, ever, and no acceptable data loss.
Their internal team was four infrastructure engineers who had run the colo competently for years and had never done a cloud migration. They were candid that they didn't know what they didn't know — a better starting position than most.
Decision path
The agreed model was embedded: Envion led the migration architecture and cutover planning, the client's engineers did the majority of the execution, and ownership was deliberately transferred so they'd run it alone afterwards. Their CTO's condition — which we liked — was that they didn't want a consultancy they'd still be paying in year two.
Months 1–2, inventory and honesty. Every workload catalogued and, more importantly, every undocumented dependency. This is where migrations die. The exercise found a scheduled job on a server nobody could account for that generated a monthly regulatory report; a hardcoded IP address in a customs authority integration that the authority had whitelisted years earlier and would need thirty days' notice to change; a licence server tied to a physical MAC address; and 400GB of archived declaration data under a seven-year retention obligation that existed on exactly one disk with no backup. That last one was fixed in week two, before any migration work, because it was a live risk independent of the project.
Month 2, sequencing by reversibility. The reversible things moved first — the ones where rollback takes minutes if something surprises you. Internal tools, reporting, batch processing. The customs integrations moved last, when the team had months of practice and confidence.
Envion contribution
Months 3–5, move and rehearse. Workloads migrated in waves, with the database replicating continuously to the cloud target throughout so cutover would be a promotion rather than a restore. The cutover procedure was run four times against production-scale data. The first rehearsal took 3 hours 40 minutes. The fourth took 51 minutes. The client's engineers led rehearsals three and four. That was the point.
Month 6, the customs integrations. Coordinated with two authorities on IP whitelisting, and ran parallel submission — the same declarations filed from both environments and reconciled — for three weeks before switching. Slow, unglamorous, and the reason nothing went wrong.
Month 7, cutover. Sunday, during the widest filing gap in the week. 47 minutes. Rollback plan tested and unused. Their team ran it; Envion was on the call and said almost nothing — the correct outcome for the last day of an embedded engagement.
Delivery
The engagement was deliberately structured to end. Envion's involvement tapered as the client's engineers took over rehearsals, then the cutover, then operations. There was no retainer, no handover document nobody reads — the capability stayed with the team because they had done the work themselves, with Envion watching rather than driving.
Outcome and evidence
Delivered in month 7 against a hard 9-month deadline. The cutover used 47 minutes of a 2-hour window; zero declarations were delayed or missed; zero data was lost. Infrastructure cost fell from €38K/month (colo plus hardware amortization) to €26K/month. Provisioning a new environment went from 3 weeks to 25 minutes, deployment from monthly and manual to daily and automated, and the previously unbacked-up 400GB regulatory archive is now protected.
Ongoing Envion involvement: none, by design.
| Metric | Before | After |
|---|---|---|
| Cutover window used | 2h available | 47 min |
| Declarations delayed or missed | — | 0 |
| Data loss | — | 0 |
| Deadline | Hard, 9 months | Delivered month 7 |
| Infrastructure cost | €38K/mo (colo + hardware) | €26K/mo |
| Provisioning a new environment | 3 weeks | 25 min |
| Recovery time objective | Undefined | 45 min (rehearsed) |
| Deployment frequency | Monthly, manual | Daily, automated |
| Unbacked-up regulatory archive | 400GB | 0 |
| Ongoing Envion involvement | — | None (by design) |
Client feedback
What the client says about this engagement
“We had a deadline we couldn't move and a team that had never done this. What I'd tell anyone in that position is to insist on the embedded model. Alex could have run the whole migration himself faster — he said so — and instead our engineers led the last two rehearsals and the actual cutover while he sat on the call quietly. That's why we're operating it ourselves now with nobody on retainer.
The undocumented-dependency phase was the part I'd underestimated. A scheduled job on a machine nobody could identify, producing a regulatory report. We'd have found that in month two of running in the cloud, when it stopped.”
From the engagement lead
What I’d tell anyone considering this
“If you have a fixed deadline and no room for data loss, the sequencing rule is simple: move the reversible things first. You want your team's first ten migrations to be the ones where a mistake costs an hour, so that by the time they touch the thing that can't fail, they've done it ten times.
Rehearse the cutover until it's boring. Four times isn't excessive; the first rehearsal will find things no amount of planning would have, and the fourth is where you learn your real number. If you've never rehearsed, your cutover estimate is fiction.
Budget serious time for undocumented dependencies. Every environment older than about three years has jobs, integrations, licences and hardcoded assumptions that exist in nobody's documentation and one person's memory. Two months on inventory sounds slow and it's the cheapest insurance in the project. And if you're hiring help for a migration, ask explicitly who will be operating the result — a consultancy that runs the whole thing and leaves gives you a working system and no capability.”
Evidence gate. This page publishes only what Envion's project records and client disclosure permissions support. Outcomes are added once verified against a baseline, a measurement period, and an approved source.
FAQ
Questions about this case
Facing a similar challenge?
Fixed deadline, no room for data loss? Envion embeds with your team, moves the reversible things first, rehearses the cutover until it’s boring — and leaves you running it alone.
Discuss a Similar ChallengeKeep exploring
Similar case studies
Executive Technology Leadership
Support for high-stakes product and AI decisions
Bring senior technology leadership into the business when the roadmap is unclear, delivery is at risk, an AI initiative needs stronger ownership, or the company needs an experienced technical voice before hiring a permanent CTO.
Discuss Interim CTO SupportCore responsibilities
- Align product and technology priorities with business goals and measurable outcomes.
- Review architecture, delivery risks, data foundations, security needs, and AI readiness.
- Lead internal teams and external partners through a practical execution plan.
- Clarify team structure, ownership, decision rights, and delivery cadence.
- Support investor, board, partner, and due-diligence conversations with credible technical judgment.
New experience
Prompt-to-Page — try it right here
Describe the landing page you want, in your own words. We turn it into a finished page and email you a private link in 5–10 minutes — no briefs, no calls, $0 to see the result.
- Describe what you want to create.
- We structure, write, and compose the page.
- You receive a private link when it is ready.
Start with a sentence — the interactive builder takes it from there.
Generate My PageSafe, respectful content only. No obligation.