Last updated
Refreshing network equipment at one site is a maintenance task. Refreshing it at hundreds of sites is a logistics program. The engineering is usually settled early: the new switch, firewall, and access point models and a standard configuration. What decides whether the program finishes on time is inventory accuracy, scheduling, staffing, and a cutover process that works the same way at every location.
Start with an accurate inventory
Every later step depends on knowing what is at each site. Pull what you can from monitoring and management tools, then confirm it on site for locations where records are old or incomplete.
- Device models, serial numbers, and locations
- Port counts in use, and what is connected to each port
- Rack space, power, and UPS capacity in each closet
- Circuit and carrier hand-off details
- Photos of each closet and rack
Sites with unknown or unusual closets are the ones that fail on cutover night. A pre-refresh site survey finds them while there is still time to plan.
Standardize the design and pre-stage
Define a small number of standard site designs, for example small, medium, and large, each with its own equipment list and port map template. Configure or pre-stage devices before they ship, using zero-touch provisioning where your platform supports it. The less configuration that happens on site, the shorter and more predictable each cutover becomes.
Pilot before you scale
Choose pilot sites that represent each standard design, including at least one difficult site. Run the full process: shipping, technician briefing, cutover, validation, and closeout. Record how long each step took and every question that came up, then update the runbook. Repeat with a second, slightly larger wave before committing to the full schedule.
Schedule in waves
Group sites into waves that can be supported at the same time. A practical wave plan considers:
- Change windows for each site type, such as after close for retail or weekends for offices
- Time zones, so remote engineers are not covering windows across too many hours
- Blackout periods, such as holiday trading, fiscal year end, or school terms
- Geography, so technicians and spares can be positioned efficiently
- Shipping lead times, so equipment arrives several days before the window
- Buffer capacity for sites that need a return visit
Assign and brief the technicians
Match technicians to the work. A switch swap with a remote engineer on the call needs someone comfortable with console sessions and port-by-port moves. An access point replacement needs someone comfortable on a ladder or lift with a clear mounting plan. Brief every technician before their first site with a short pack:
- The runbook for the site design
- The port map and photos of the existing closet
- The check-in and check-out procedure
- The escalation contact and bridge line
- The closeout shot list and deliverables
Plan remote engineer coverage at the same time. One engineer can support several concurrent cutovers only if the runbook is reliable. During early waves, keep the ratio low.
Write a cutover runbook
- Pre-checks: confirm equipment on site, power available, existing configuration backed up, and the remote engineer on the call.
- Photograph the existing rack and label every cable in use.
- Install and power the new device, and confirm management access.
- Move cables port by port to the port map.
- Validate: links up, critical systems such as POS or phones tested, internet and failover confirmed.
- Decide go or rollback at a set time within the window.
- Remove and label the old device for return.
- Photograph the finished rack and close out.
Rollback criteria must be written before the window, not decided during it. Define what failure looks like, the latest time a rollback can start and still finish inside the window, and who makes the call.
Report every day
A daily status report keeps everyone working from the same picture: sites completed, rolled back, or rescheduled, issues and their owners, and the next day's schedule. Track return visits and their causes, because a cause that repeats points to a gap in the runbook, the shipping process, or the survey data.
Close out each site and the program
Each site should close out with before and after photos, the confirmed port map, serial numbers in and out, validation results, and engineer sign-off. At the end of the program, combine them into a final report with completed sites, exceptions and owners, and returned equipment. That record becomes the baseline for the next refresh.
Common causes of slipped schedules
- Equipment shipped late or to the wrong address
- Closets that do not match the inventory
- Technicians briefed on the morning of the cutover
- No agreed rollback point, so windows overrun
- Old equipment left on site with no return plan
Relion supports multi-site refreshes with briefed technicians, daily status reporting, and a per-site closeout package. See IT Projects & Rollouts and our network equipment service.