Skip to content

Use Cases

Three common deployment topologies: a centralized regional egress gateway for teams, composition with an existing corporate VPN stack, and a personal self-hosted network for home routers and workstations.

Scenario 1: Centralized Regional Egress for Teams

An engineering or corporate environment where partner APIs, B2B portals, cloud registries, or QA staging environments expect traffic from specific geographic regions or static egress IPs while the team operates elsewhere.

                        [office / cloud workloads]
                                   │
                                   ▼
               ┌───────────────────────────────────────┐
               │  Relay master + edge + exit           │
               │  on the organization's main gateway.  │
               └───────────────────┬───────────────────┘
                                   │
                   ┌───────────────┴───────────────┐
                   │                               │
            Route A:                        Default:
            domains & CIDRs of              traffic exits directly
            partner services in Region X →  from the local gateway.
            exit node in Region X.
                   │
                   ▼
            ┌──────────────────────┐
            │ Relay exit node      │
            │ in Region X          │
            └──────────┬───────────┘
                       │
                       ▼
            [geo-sensitive / IP-allowlisted services]

Key Benefits

  • Predictable regional egress for partner integrations and services that filter or localize responses by source IP.
  • Policy-level high availability: list multiple exit nodes on any route; if packet loss exceeds 20% or a link drops, the gateway fails over automatically to the backup node and lazily returns to primary once healthy. If every allowed exit node is unreachable, connections fail cleanly rather than leaking from the local WAN IP.
  • Per-subnet & CI split routing: assign dedicated exit regions and isolated routing tables (client_defaults) to specific VLANs, QA subnets, or build runners without spinning up separate gateways.
  • Zero client configuration: changing an egress region is a single policy update — developer laptops and application containers stay untouched.
  • Data-plane continuity without the master: nodes cache policy.snapshot locally and keep forwarding traffic even if the control plane is offline for maintenance.

Minimal Topology

Two nodes: a gateway node (master + edge + exit) + one regional exit VM. Scale horizontally by adding lightweight VMs in additional regions and registering them in policy.

Scenario 2: Layering Behind an Existing Corporate VPN

When your organization already runs an established remote-access VPN such as OpenVPN managed via PritunlSolid open-source OpenVPN management layer — saves hours compared to manual PKI and profile management. , you do not need to touch user clients to add multi-region policy routing.

[remote employees]
      │
      │ OpenVPN client
      ▼
[corporate VPN gateway]
   ├── OpenVPN / Pritunl server
   │       terminates user sessions (MFA, SSO, audit),
   │       delivers traffic onto tun0.
   │
   └── ololo-relay (capture_iifs: [tun0])
           │
           │ applies SNI / CIDR / per-client routing policy
           ▼
        ┌──┴────────────────────────────────────┐
        │                                       │
   exits directly from                   forwarded over TLS mesh
   the local gateway                     to a regional relay exit
   (local exit role)                            │
                                                ▼
                                        [regional relay exit]
                                                │
                                                ▼
                                        [egress in target region]

Key Benefits

  • Zero disruption to existing infrastructure. Employees keep using their existing OpenVPN profiles, 2FA, and SSO.
  • Clean separation of responsibilities. OpenVPN/Pritunl handles user authentication, OTP, and connection auditing; ololo-relay handles L3 policy routing, SNI extraction, automatic MSS clamping (TCPMSS 1280), and regional failover.
  • Localization & multi-region QA. Verify how your own product renders across regions — currencies, pricing tiers, regional payment providers, and geo-dependent content — right from your standard work VPN.

Scenario 3: Personal Self-Hosted Mesh, Home Router & Workstations

A personal homelab setup combining one or two small cloud VMs with a home router (Keenetic / OpenWrt / Linux gateway) and personal workstations or laptops on the road.

[home router / LAN segment]        [laptop on the road (OpenVPN / WG)]
              │                                      │
              ▼                                      ▼
      ┌─────────────────────────────────────────────────────┐
      │  Relay ingress node (edge + master)                 │
      │  unified domain/CIDR routes & per-device rules      │
      └───────────────┬──────────────────────┬──────────────┘
                      │                      │
         Local & domestic subnets →    Cloud lab & external work tools →
         direct egress                 exit node in selected region

Key Benefits for Individuals

  • 100% self-hosted ownership. Every binary, key, and routing rule lives on your own VMs — no third-party accounts, shared exit pools, or opaque cloud controllers.
  • Set-and-forget home network. Attach a home router or dedicated Wi-Fi SSID once; laptops, workstations, and lab boxes get policy routing automatically without running background agents on every machine.
  • Consistent policy on the road. When traveling, connect a laptop to your home or cloud gateway to keep encrypted transport and the exact same local-bypass and regional routing rules.
  • Per-device rules & torrent guard. Assign custom gateway rules to specific home devices (such as a work laptop vs. smart TV) from the desktop GUI, while built-in block_bittorrent prevents a forgotten torrent client at home from getting your cloud VPS suspended.