It is peak hour, and customers are queuing to pay. Suddenly, the internet connection drops: checkout terminals cannot access the central system, staff cannot retrieve promotions, and managers repeatedly call IT for support.
At a factory, a similar disruption may prevent the warehouse from receiving dispatch orders or delay production data updates to headquarters.
Not every activity stops immediately when the internet goes down. But businesses need to know exactly which operations can continue, which will be affected, and for how long.
Even with two internet connections, an offline operating plan is essential
A backup connection helps reduce the risk of disruption. However, businesses must also account for network equipment failures, power outages at individual sites, and problems affecting central services.
Alongside asking, “Is our internet connection reliable?”, businesses should ask: “If we cannot connect to the central system, which tasks can we still perform locally?”
Bring processing closer to where work happens
Edge computing processes data on devices or servers close to where it is generated. With an appropriate design, selected tasks can run locally at a store, warehouse, or factory, while data is temporarily stored and transmitted to the central system once connectivity returns.
In practice, businesses can consider:
- Supermarkets: POS terminals use previously synchronized product and pricing data to record transactions within defined operating limits.
- Warehouses: Staff continue processing downloaded orders and retain activity records for later synchronization.
- Factories: Local equipment and applications continue performing functions designed to operate independently.
- AI cameras: Selected analytics run on devices or local servers when cloud access is unavailable.
These capabilities must be designed and tested for each application. Having an on-site server alone does not guarantee that all business processes will work without internet access. Electronic payments, account authentication, and online data lookups may still depend on external systems.
When connectivity returns, will the data reconcile?
Keeping operations running offline is only half the challenge. The other half is transferring data back to the central system completely, in the correct sequence, and without duplicate records.
For example, sales recorded at checkout must be reconciled with the central system, while warehouse dispatch records must update without creating inventory discrepancies. Temporary storage capacity and data retention periods also require planning. Microsoft’s documentation on offline edge operations explains that queued data storage depends on configured retention settings and available disk space.
Five questions to include in deployment requirements
- Which business processes must continue during a connectivity outage?
- How long must they operate, and how many transactions or devices must be supported?
- Which data must be available locally, and who can access it?
- How will the system synchronize data and resolve discrepancies once connectivity returns?
- Has an actual disconnection drill been conducted, including recovery and reconciliation?
The answers should become verifiable acceptance criteria.
Metatek helps build infrastructure around operational needs
Metatek helps businesses review dependencies between applications, connectivity, and equipment; identify critical processes; and develop an appropriate combination of local processing, cloud services, and backup connectivity.
The intended value is reduced downtime, fewer missing transaction records, and clear procedures that staff can follow when disruptions occur.
