Recovering a Total ERP Database Outage in a Single Morning
The Challenge
This client is a metrology and calibration company that runs quoting, invoicing and accounting on the Spire business management platform, with its data in a PostgreSQL database on an on-premise server. One morning in June 2026, every workstation in the office lost its connection to that database at the same moment. Nobody could open Spire. For a business whose day is built around issuing certificates, quotes and invoices, that is not an inconvenience — it is the company standing still.
There was no obvious cause to work from. The server was powered on. The network was up. Spire itself had not been touched, updated or moved. The symptom pointed at the application; the fault was somewhere else entirely.
What We Did
- Took the call that morning and worked the problem remotely from symptom to root cause in a single session, rather than restarting services and hoping.
- Found the real cause: the server’s network profile had changed, and the new profile applied a stricter firewall posture that blocked PostgreSQL connections. That is precisely why every workstation failed simultaneously while everything else on the network looked perfectly healthy.
- Corrected the profile and restored database connectivity to all three affected workstations, with Spire coming back up with no data loss.
- Delivered a written incident report the same morning — sent at 10:13 a.m. — covering what failed, why it failed, what was done about it, and what it means going forward.
- Followed up three days later with two written proposals: one to design, configure and validate a proper backup solution for the Spire accounting database, and one to set up secure, password-protected external file sharing inside the client’s existing Microsoft 365 tenant so files leave the company through controlled links instead of email attachments.
- Continue to manage this client’s Microsoft 365 subscriptions and licensing, as we have since their original Spire installation support, their workstation refresh, and their move to OneDrive.
The Result
Service was fully restored the same morning it failed, with no data loss and a documented root cause — so the failure mode is closed rather than waiting to recur. The client had a written incident report in hand before the morning was out, and a backup and file-sharing hardening plan on the table within three days. An outage became a resilience upgrade.
Related Services & Guides
Facing something similar in your business?
We handle engagements like this every week for businesses across the GTA — from email security audits to full Microsoft 365 management.