SD-WAN Deployment, Site by Site, Across All 48 States.
You already picked the platform. What you need now is hands: someone to survey every location, stage the edge appliances, chase the carrier, mount the gear, cut the traffic over inside a window that works for that site, and prove it came up.
SRS Networks executes SD-WAN rollouts in the field — 20 sites or 500 — with appliances pre-staged at our West and East Coast facilities, circuits coordinated per site, and every turn-up validated and documented before we leave.
SD-WAN deployment is the field execution half of an SD-WAN program: site surveys, edge appliance staging and configuration, shipping and install sequencing, carrier circuit coordination, scheduled cutover windows, per-site turn-up validation, rollback planning, and legacy hardware decommission. The design and the platform decision sit upstream of it. This page is about getting the boxes on walls and the traffic moved, in every location, on a schedule that holds.
SRS Networks is a nationwide network deployment contractor headquartered in Salinas, California, executing multi-site SD-WAN rollouts for retail chains, healthcare systems, restaurant groups, and distributed enterprises across all 48 contiguous states since 1996. We have completed 500+ deployments across 5,000+ sites from five offices with in-house W-2 field leads, West and East Coast staging facilities, and every site tracked live in our Project Command Center. If you are still choosing an architecture rather than installing one, start with our SD-WAN and multi-site networking overview and come back here when the estate is ready to roll.
SD-WAN Rollouts Fail in the Field, Not in the Design
The architecture is rarely the thing that kills a multi-site rollout. Logistics, carriers, and scheduling are. Four patterns account for most of the schedule slip we get called in to fix.
300 Appliances Drop-Shipped to 300 Addresses
Boxes land at store back doors with no receiving process, get signed for by whoever is closest, and go missing. At scale, roughly one in six shipments is wrong, damaged, or gone by install day — and the tech finds out standing in the parking lot.
Carrier Ready Dates That Move
The broadband or DIA order slips two weeks and nobody tells the field. The tech is already dispatched, arrives to a dead demarc, and you pay for a truck roll that installed nothing. Multiply by a few dozen sites and the schedule stops meaning anything.
Cutover Windows Across Four Time Zones
A store cannot cut over at 2pm, a clinic cannot cut over mid-session, and a distribution center runs three shifts. Every site has its own window, and coordinating hundreds of them by email is how rollouts stall at site 40.
No Rollback When the Tunnel Drops at 3am
The tech left at midnight, the legacy edge is already in a box on a truck, and the overlay flaps four hours later. Without a written rollback path and the old gear still racked, that site is dark until the next dispatch lands.
A Repeatable Per-Site Deployment Runbook.
Site 1 and site 300 get the same sequence, the same kit contents, the same validation, and the same closeout record. That is what makes a large estate predictable. It runs alongside our multi-site network deployment and internet provider coordination work when the circuits are changing too.
Six Phases, Repeated at Every Location
This is the order we run a site in, and the order we report against. Nothing moves to the next phase until the previous one is evidenced.
Survey Every Site
We document the demarc type and location, the existing router or firewall make and model, the RU or wall space available, the outlets and UPS capacity, the cable path from demarc to rack, and the LTE or 5G signal at the exact spot the antenna will sit. Sites that need remediation get flagged now, not on install day.
Stage and Kit the Edge
Appliances land at our West or East Coast facility. Each one is serialized to a site ID, licensed, claimed to your orchestrator, brought to the standard firmware, configured from your template, SIM installed and activated, port-labeled, and burned in on the bench before it is boxed with power supply, mounting hardware, patch cords, optics, antenna, and a site-specific install sheet.
Coordinate the Circuits
We hold the carrier ticket per site — order, FOC date, handoff type, and speed — and we test the circuit before cutover night rather than discovering a dead demarc with a tech on site. LTE or 5G backup gets activated and speed-checked at the real mount location, not at the loading dock.
Ship and Sequence
Kits ship to land two to three days ahead of the window, addressed to a named receiver who knows it is coming. Sites are sequenced by region and by circuit readiness so crews cluster geographically instead of criss-crossing states, and a slipped circuit re-sequences the site rather than burning the dispatch.
Cut Over and Turn Up
Inside the site's window we cut traffic, confirm tunnels to your hubs and gateways establish, confirm the routing and QoS policy pushed from the controller, verify DHCP, DNS, and VLAN behavior, then test the applications that actually matter at that site — POS, EMR, UCaaS, ERP. We pull the primary transport to force a failover, confirm sessions survive, and fail back.
Document, Soak, Decommission
Closeout captures serials, circuit IDs, WAN addressing, SIM identifiers, throughput and failover results, config version, and install photos. The legacy edge stays racked and reversible through an agreed soak period so a 3am problem has a rollback instead of an outage. After the soak we return, pull the old gear, and hand it to your reverse-logistics process.
Scope of an SD-WAN Deployment Engagement
Four workstreams cover the estate from first survey through the day the last legacy router leaves the building.
Site Survey & Readiness
Before anyone is dispatched we document each location — demarc type and location, existing edge hardware, rack or wall space, available power, and the LTE read at the intended mount point.
Staging, Kitting & Config
Gear ships to our staging facility, not to your sites. Each unit is serialized to a site ID, licensed, claimed to your orchestrator, firmware-matched, burned in, and boxed as a complete site kit.
Install, Cutover & Turn-Up
Crews mount, cable, and power the edge, bring the transports up, and hold traffic until the window opens. Cutover runs to the site's hours with a rollback path written before the first cable moves.
Validation, Docs & Decommission
Every site closes with evidence — serials, circuit IDs, WAN addressing, throughput and failover results, install photos. Legacy edge stays reversible through a soak period, then gets pulled and returned.
You Picked the Vendor. We Install It.
We are the hands, not the resale channel. Whatever platform your team standardized on, we stage to your tenant, install to your design, and validate against your policy.
VMware VeloCloud
Edges activated against your Orchestrator with the profile already assigned, so the unit calls home configured. Business policy and link steering verified per site at turn-up.
Cisco Meraki MX
Appliances claimed into your dashboard org and network before shipping, templates bound, and AutoVPN hub connectivity confirmed on site before traffic moves.
Fortinet Secure SD-WAN
FortiGates staged, firmware-matched, and registered to your FortiManager or FortiCloud instance. SD-WAN rules, SLA probes, and failover behavior validated at cutover.
Cato, Prisma, EdgeConnect
Cato Sockets, Palo Alto Prisma ION, Aruba EdgeConnect, Versa, and Catalyst SD-WAN all get the same treatment — staged to your tenant, installed to your design, validated the same way.
Rollouts Break on Logistics. That's What We Own.
Anyone can bolt an appliance to a wall. Doing it at 300 addresses on a schedule, with carriers slipping and every site keeping different hours, is a logistics discipline. We built the staging facilities and the PM structure around that problem.
What a 300-site SD-WAN rollout actually looks like.
Say you have 300 stores across 26 states, the platform decision is made, the purchase order is cut, and your leadership has committed to a completion date. The first thing that happens is not an install. It is a survey pass. We need to know which sites have a fiber handoff and which still terminate on coax, which have two free RU and which have a plywood board and a surge strip, which have an outlet within reach of the mount, and what the LTE read is at the spot the antenna will actually sit — not at the front counter where the phone gets five bars. That pass is what turns a 300-site date into a 300-site plan, because it tells you which twenty locations need electrical or mounting remediation before a network tech is worth dispatching.
Next, the gear does not go to your stores. It comes to our West or East Coast staging facility. Each appliance gets serialized against a site ID, licensed, claimed into your orchestrator with the right template bound, brought to the firmware version you standardized on, SIM installed and activated, port-labeled, and burned in on the bench. Then it gets boxed as a complete kit — appliance, power supply, mounting hardware, patch cords, optics, antenna, and a one-page install sheet specific to that store. This is the single highest-value step in the whole program. Drop-shipping from the vendor to 300 back doors is where rollouts go to die: at scale roughly one in six shipments arrives damaged, wrong, opened, or simply cannot be found, and every one of those is a wasted truck roll.
In parallel we hold the carrier. If new broadband or DIA is going in, we own the ticket per site — order placed, FOC date tracked, handoff type and speed confirmed, circuit tested before cutover night. Carriers slip. That is not a surprise, it is a planning input. When a site slips, we re-sequence it into a later cluster instead of dispatching a tech to a dead demarc. Kits ship to land two or three days ahead of the window, addressed to a named receiver who knows it is coming, and crews get routed regionally so a Tuesday in the Dallas cluster is six sites, not one site and four hours of highway.
Cutover runs on the site's clock, not ours — pre-open for retail, after last patient for clinics, third-shift changeover for distribution. We cut traffic, confirm the tunnels to your hubs come up, confirm the routing and QoS policy actually pushed from the controller rather than assuming it did, check DHCP, DNS, and VLAN behavior, and then test the applications that matter at that site: the POS talks to the payment processor, the EMR loads, the phones register, the ERP client stays up. Then we pull the primary transport on purpose to force a failover, watch sessions survive, and fail back. A site that has never failed over is a site whose failover does not work yet.
Nobody leaves with a handshake. Closeout per site is serials, circuit IDs and carrier account, WAN addressing, SIM identifiers, throughput and failover results, config version, and photos of the install. The legacy edge stays racked and reversible through an agreed soak period, so a problem at 3am on night four has a rollback path instead of an outage. After the soak we come back, pull the old gear, and hand it to your reverse-logistics or our network decommissioning process. Throughout, you watch site status roll up live in the Project Command Center instead of asking for a spreadsheet on Fridays. One PM, one standard, one PO across all 48 states.
Explore More from SRS Networks
SD-WAN Deployment FAQs
The questions IT directors and network managers ask us before handing over a multi-site SD-WAN rollout.
Send Us the Site List. We'll Send Back a Schedule.
Give us the locations, the platform, and the date you have to hit. You get a survey plan, a staging and shipping sequence, a cutover calendar built around each site's hours, and one project manager who owns all of it across 48 states.