By Impress IT Solutions | Houston, TX
—
When a Houston-area construction company reached out to us, the situation wasn’t ideal. Their previous IT provider had hosted their entire infrastructure on AWS, and the transition to a new provider came with minimal notice. There was no extended runway, no luxury of a slow rollover — just a tight deadline and a business that absolutely could not afford to go dark.
Eight weeks later, the company was fully operational on Microsoft Azure, every application had been migrated, 16 field sites were connected, and the old AWS environment had been completely decommissioned. Not a single minute of unplanned downtime.
Here’s how we did it.
—
The Challenge
Construction companies live and die by uptime. When your project management software goes down, crews stop. When accounting systems go offline, invoices freeze. This company ran a complex stack of applications — including Oracle P6 Primavera for project scheduling, PlanSwift for takeoffs, Vista Viewpoint for accounting, GCPay for subcontractor payments, ACT! for CRM, and a collection of legacy Access databases that had grown organically over years.
Every one of them had its own migration requirements. Some were cloud-hosted SaaS tools that just needed IP whitelist updates with vendors. Others required spinning up SQL servers, importing databases, and configuring application layers from scratch.
Oh — and we also had 16 field sites spread across the region, including one location that needed a Starlink connection put into passthrough mode to feed a Cradlepoint router.
—
The Strategy: Build First, Cut Over Fast
The key to zero-downtime migration is simple in concept but demanding in execution: never touch a production system until the replacement is fully tested and ready.
We spent the first 30 days (August 1–30) building the entire Azure environment in parallel with the live AWS setup. This included:
Spinning up the Azure tenancy through a Nerdio-assisted deployment
Establishing an AWS-to-Azure site-to-site VPN so both environments could communicate during the transition
Deploying and configuring every application server — SQL instances for P6 Primavera and ACT!, Oracle web server for P6 user management, print server for 6 office printers
Deploying Appgate — a zero-trust network access solution — connected via SAML to Azure Active Directory, replacing legacy VPN
Configuring all 16 Cradlepoint devices with site-to-site VPN tunnels back to the new Azure environment
Only once every piece was tested did we begin cutting users over.
—
The 5-Day Cutover Sprint
Starting August 31st, we executed a carefully sequenced user migration:
Day 1 (Aug 31) — P6 Primavera
We performed a fresh overnight backup of the live SQL database and imported it into the new Azure SQL server. At 7am, we were on-site to switch corporate users to the new server. Remote users followed the same day.
Days 2–3 (Sep 1–2) — PlanSwift, Printing & Scanning
PlanSwift gave us our biggest curveball. Initial testing on Azure cloud storage was too slow — users complained immediately. We pivoted to a hybrid NAS architecture: a local NAS drive hosting the PlanSwift data with Azure as the backup layer. Problem solved. We also stood up a print server on the Azure domain controller and connected all 6 office printers, then migrated the document scanning workflow from the old server to Egnyte with folder sync and address book updates.
Day 3 (Sep 3) — ACT! & Access Databases
A second SQL server was deployed specifically for ACT! We imported the ACT! database backup, transferred the legacy Access databases, and performed an early-morning on-site visit to cut users over. The Access accounting database had a complication during migration — we resolved it on-site and confirmed functionality with the client’s controller.
Days 3–4 — GCPay Connector
GCPay required coordination with Vista Viewpoint to whitelist the new Azure server’s IP address. We submitted the service request, waited for confirmation, then ran a second GCPay meeting to complete the migration and test it with the accounting team.
Day 4 (Sep 3–4) — Acronis Backup
All production servers were enrolled in Acronis backup: the domain controllers, the RDS server, the SQL servers, and the file server.
Day 5 (Sep 4) — DNS, FSMO & AWS Shutdown
We reconfigured the WatchGuard firewall with updated DNS settings, updated all printers with new DNS, seized the FSMO roles to the new Azure domain controller, and demoted the legacy DC to a member server. Then — one by one — we shut down the remaining AWS servers.
Migration complete.
—
What We Learned
A few takeaways worth sharing for any IT professional or construction company considering a similar move:
Cloud storage isn’t always faster than local. For applications like PlanSwift that depend on frequent small file reads, a hybrid NAS approach often outperforms pure cloud storage.
Undocumented dependencies will always surface. When we shut down one of the AWS servers, we discovered a user had a mapped network drive pointing to it. Always audit active connections before decommissioning anything.
Third-party vendor lead times are your critical path. GCPay’s IP whitelist request through Vista Viewpoint had to be submitted, approved, and confirmed before migration could happen. Build vendor coordination time into your timeline.
Zero-downtime requires a parallel environment. There’s no shortcut. If you want zero downtime, you need to build the replacement before you touch the original.
Early-morning cutover visits matter. Cutting over users before the business day starts means that by 9am, everything is working and the day proceeds normally.
—
Is Your Construction Business Ready for Azure?
Whether you’re moving away from a legacy provider, modernizing aging infrastructure, or simply looking for better reliability and security, Microsoft Azure offers construction companies a powerful, compliant, and scalable platform.
At Impress IT Solutions, we’ve spent over 23 years supporting construction companies, contractors, and field operations teams across the Houston area. We understand the software you use, the connectivity challenges you face, and the cost of downtime on an active project.
Download our free white paper for the complete technical walkthrough of this project — including the full timeline, risk management approach, lessons learned, and application-by-application migration details.
👉 Download the Free White Paper
Or, if you’d like to talk about your own infrastructure, book a free discovery call with our team.
📞 281-944-3983 | ✉️ info@impresscomputers.com | 🌐 impresscomputers.com
—
Impress IT Solutions is a managed IT services provider based in Katy, TX, serving Houston-area businesses in construction, engineering, finance, law, and manufacturing.
