One overlay across all your sites — with the QoS classes you actually need.
L3VPN (RFC 4364) and VPLS / EVPN VPNs over our Punjab MPLS core. Any-to-any reachability between branch offices, data centres and partner DMZs, with six QoS classes and per-site traffic-engineering controls.
A private network that scales past two sites.
When you have three or more sites that need to talk to each other, point-to-point circuits stop being sensible — every new site means a new circuit, a new routing decision, and a fresh fight with the firewall team. An MPLS VPN solves that by giving you a single overlay where every site is one BGP hop away from every other, with route distribution handled by us and policy controlled by you.
On our network, L3VPNs run as RFC 4364 BGP/MPLS VPNs with route targets and route distinguishers per customer. L2VPNs are delivered as VPLS (RFC 4761/4762) for multipoint, PWE3 for point-to-point, and EVPN where customers are looking for active / active MAC-mobility. Six QoS classes are available end-to-end, and per-site bandwidth profiles let you size the access link independently of the VPN's total reach.
The kind of networks that live on this.
Bank branch network
30–300 branches across Punjab on a single L3VPN, with a strict EF class for ATM transactions and a separate AF for video.
Multi-DC enterprise core
Two production DCs, one DR, and a handful of regional offices on a hub-and-spoke L3VPN with VRF leaks where needed.
Government / public-sector network
Provincial agencies on a closed VPN with strict route-target isolation and a centralised, audited internet break-out.
Voice carrier overlay
VoIP carrier with media gateways across Lahore, Multan and Islamabad, on a VPLS instance with PDV-controlled EF class.
Wholesale carrier extending reach
Other ISPs/LDIs riding our MPLS to extend their own VPN service into Punjab without separate point-to-point builds.
Operational tech (OT) network
Manufacturing operator separating IT and OT traffic on different VRFs, with no inter-VRF route leakage by policy.
Service options and SLA.
| Metric | Target | Notes |
|---|---|---|
| L3VPN | RFC 4364 BGP/MPLS | Any-to-any, hub-spoke, partial mesh |
| L2VPN multipoint | VPLS (RFC 4761/4762) | EVPN-VPLS on request |
| L2VPN point-to-point | PWE3 / EVPN-VPWS | Same as DPLC underneath |
| Routing protocol | Static / BGP / OSPF | Per VRF; we run the PE-CE side |
| QoS classes | 6 (EF, AF4x, AF3x, AF2x, AF1x, BE) | DSCP / 802.1p preserved |
| Multicast | mLDP / MVPN supported | PIM-SSM / SM both available |
| Metric | Target | Notes |
|---|---|---|
| Availability per site | 99.95% | Single-feed; 99.99% on dual-PE |
| Intra-VPN p95 latency | Per route | Published in service description |
| Packet delay variation | ≤ 3 ms p99 (EF) | Real-time class only |
| Packet loss | ≤ 1 × 10⁻⁴ (EF/AF) | Within CIR |
| Site addition | 10 – 14 days on-net | Off-net depends on civils |
| Reporting | Customer portal | Per-site, per-class graphs |
Bringing an MPLS VPN up.
- 01
Topology design
Day 0 – 7. Sites, capacities, routing protocol, QoS budget. We deliver an LLD before signing.
- 02
PE config + access
Day 7 – 25. VRFs, route targets, access ports per site. CE devices brought up if managed.
- 03
Routing burn-in
Day 25 – 32. BGP / OSPF stabilised, advertised prefixes verified, QoS marking checked end-to-end.
- 04
Cutover + monitoring
Day 32 – 40. Live cutover per site (rollback ready), portal access handed over, monthly SLA review scheduled.
Pre-sales questions we hear.
When do I want L3VPN vs VPLS?
Do you offer managed CE?
What internet break-out options do you offer?
Can the VPN extend off-net?
How is QoS enforced?
What about IPv6?
Do you support route reflection or partial mesh?
A list of sites and a topology preference is enough to start.
Tell us how many sites, the rough capacity per site and what your QoS picture looks like. We'll come back with an LLD, a price, and a per-site latency table.