Network-as-a-Service (NaaS) & Network-on-Demand
Consumption-based enterprise connectivity you control from a portal or API - turn bandwidth up or down in minutes, create virtual cross-connects between data centres, clouds and PoPs, and pay only for what you use.
< 5 min
Bandwidth change, portal or API
40+
On-net PoPs and cloud on-ramps
1.4 Tbps
Backbone capacity
What is Network-as-a-Service (NaaS) & Network-on-Demand?
Dollu Network-as-a-Service (NaaS) replaces the fixed-term, fixed-bandwidth circuit with connectivity you consume like cloud. Once a site, data-centre cage or cloud region is on-net, every service on it - Ethernet virtual connections, internet breakout, cloud on-ramps, layer-3 VPN membership - is created, resized and torn down from the Dollu portal or REST API, with changes taking effect in minutes rather than the 30–90 days of a traditional circuit order. Bandwidth is metered and billed on what you actually use, with optional committed floors where you want the best unit price.
The service runs over Dollu's 1.4 Tbps MPLS and segment-routing backbone, 40+ points of presence and direct interconnects with the major carrier-neutral data-centre operators and public clouds. Virtual cross-connects let you link a cage in Frankfurt to a cage in Singapore, a Mumbai office to AWS ap-south-1, or a partner's network to yours at an exchange point, without waiting for a letter of authorisation and a technician with a fibre patch. The physical last-mile port is ordered once; everything above it is software, provisioned by the same orchestration our own engineers use.
Orchestration is what makes it usable at scale. Each connection is a declarative object with endpoints, bandwidth, class of service and schedule. You can burst a 1 Gbps link to 10 Gbps for a nightly replication window and drop it back automatically, template a standard site profile and roll it to 200 branches, or let your own automation call the API when a Kubernetes cluster or a new cloud region comes up. Terraform and Ansible modules, webhooks and MEF LSO Sonata- and Presto-aligned APIs are provided, and every change is logged with who, what and when.
Telemetry is streamed, not polled. Per-connection utilisation, latency, jitter, loss and availability are visible in the portal, exported through streaming telemetry using gNMI and OpenConfig, and pushed into your observability stack via Prometheus and Splunk exporters, with SLA reports generated per connection each month. The result is a network that follows workloads instead of the other way round, and a bill that tracks the business rather than a capacity forecast written five years ago.
Hybrid WAN: MPLS, internet and 5G under one SD-WAN policy.
Why choose Dollu for naas & on-demand.
The advantages of buying from a carrier that owns its network, interconnects and operations - rather than a reseller.
Capacity that tracks demand
Scale a link for a migration, a product launch or quarter-end processing and scale it back afterwards, so you stop paying year-round for the peak week.
Days become minutes
New connections between on-net locations, bandwidth changes and cloud on-ramps are live in minutes from the portal or API, with no order forms, site surveys or engineer visits.
No stranded circuits
Decommission a data centre or leave a cloud region and the connections disappear from your bill the same day; there are no early-termination penalties on on-demand services.
Automation-native
Treat the WAN like infrastructure-as-code: version-controlled Terraform, CI pipelines that provision network alongside compute, and webhooks that react to utilisation thresholds.
One fabric across DCs and clouds
Data centres, public clouds, exchange points, partners and your MPLS or SD-WAN estate share a single layer-2 and layer-3 fabric with consistent policy, monitoring and support.
Capabilities in detail.
Everything included with Network-as-a-Service (NaaS) & Network-on-Demand - the platform features, options and controls you get from day one.
- 01
On-demand Ethernet virtual connections
MEF 3.0-aligned E-Line, E-LAN and E-Tree services between any on-net ports at 10 Mbps to 100 Gbps, with per-EVC bandwidth profiles, VLAN mapping, class-of-service marking and protected or unprotected path options.
- 02
Virtual cross-connects and DC interconnect
Software-defined cross-connects inside and between carrier-neutral data centres in London, Frankfurt, Amsterdam, Ashburn, Mumbai, Singapore and partner facilities worldwide, including connections to third-party carriers and internet exchanges.
- 03
Cloud on-ramps
Private connectivity to AWS Direct Connect, Azure ExpressRoute, Google Cloud Interconnect and Oracle FastConnect at hosted-connection speeds from 50 Mbps to 10 Gbps, provisioned end to end from the Dollu portal including the cloud-side virtual interface.
- 04
Internet and VPN on demand
Add dedicated internet access or membership of a layer-3 IP VPN to any port with metered bandwidth, DDoS scrubbing options and BGP or static routing, and change the bandwidth without a new contract.
- 05
Bandwidth calendar and bursting
Schedule recurring or one-off bandwidth changes, define automatic burst rules based on utilisation thresholds, and set caps and approval workflows so finance and network teams stay in control.
- 06
Orchestration API and infrastructure-as-code
REST API with OpenAPI specification, Terraform provider, Ansible collection, webhooks for state changes and MEF LSO Sonata and Presto alignment for partner-to-partner automation, all with role-based access and audit logging.
- 07
Telemetry and analytics
Per-connection utilisation, latency, jitter, loss and availability at one-minute granularity; streaming telemetry via gNMI/OpenConfig; Prometheus, Splunk and syslog exporters; monthly SLA reports and capacity trend analysis.
- 08
Commercial models
Pure pay-as-you-go per Mbps-hour, committed-plus-burst with a discounted floor, or reserved capacity for steady-state links, mixed freely across your estate and consolidated on one invoice in USD, EUR, GBP or INR.
How it works.
From first conversation to live traffic - a tracked, engineer-led onboarding with a named owner at every step.
- Step 01
Get on-net
Order a physical port at a Dollu PoP, a supported data centre or via last-mile access at your site. This is the only step with a lead time; ports are 1G, 10G or 100G.
- Step 02
Design in the portal
Create connections between your ports, cloud regions and partners, set bandwidth, class of service and protection, and see the price before you commit. Services activate in minutes.
- Step 03
Automate
Move recurring changes into Terraform, CI pipelines or scheduled burst rules using the API, and integrate telemetry with your monitoring so the network is managed like the rest of your stack.
- Step 04
Optimise
Review utilisation and spend with your account manager, convert steady links to committed pricing and keep variable ones on demand, and retire ports and connections you no longer need.
MPLS, DIA, broadband or 4G/5G.
No single transport suits every site. Most Dollu WAN designs mix them - an SD-WAN overlay stitches the underlays together and steers voice and critical apps onto the best path.
| Dimension | MPLS / IP-VPN | Dedicated internet access | Business broadband | 4G / 5G |
|---|---|---|---|---|
| SLA | Highest - end-to-end latency, jitter, packet-loss and availability targets | High - availability and repair time on the circuit; the wider internet path is best effort | Best effort; consumer-grade repair times | Best effort; dependent on radio coverage and cell load |
| QoS | End-to-end classes of service across the Dollu core | Dedicated, symmetrical bandwidth; QoS to the network edge only | Contended and asymmetric; no QoS guarantees | Shared spectrum; private 5G adds on-site QoS and slicing |
| Cost position | Highest per Mbps | Mid | Lowest | Low - per SIM plus data |
| Lead time | Longest - new fibre and cross-connects; weeks to months | Weeks; faster where the building is already lit | Days to weeks | Days - hardware ships and activates on power-up |
| Best for | Voice-critical sites, regulated data, hub-and-spoke WANs | Head offices, data centres, cloud on-ramps, SIP trunks | Small sites, backup paths, SD-WAN underlay | Rapid deployment, temporary sites, resilience overlay, IoT and fleets |
Who uses this and why.
Typical deployments across carriers, enterprises, platforms and contact centres.
Data-centre migration and DR replication
Stand up 10–100 Gbps between old and new facilities for the migration window, then drop to a modest replication link, without a multi-year contract for either.
Multi-cloud architectures
Private, low-latency paths between AWS, Azure, Google Cloud, Oracle and your data centres, resized as workloads move and provisioned by the same pipelines that deploy them.
Media and live events
Temporary high-capacity connections between venues, production hubs and cloud media platforms for the duration of an event, with protected paths and guaranteed bandwidth.
SaaS and platform companies
Connect new customers, partners and regions on demand as the business grows, exposing connectivity through your own portal via the Dollu API.
Retail and seasonal businesses
Scale connectivity into peak trading periods and back down afterwards, keeping the base cost aligned with typical demand.
Financial services and exchanges
Deterministic, monitored connections between trading venues, market-data providers and colocation sites, added or upgraded without procurement cycles.
Technical & commercial specifications.
Key parameters at a glance. Ask us for the full service description and SLA document.
| Ports | 1G, 10G, 100G at Dollu PoPs and partner data centres; last-mile access at customer sites |
|---|---|
| Bandwidth | 10 Mbps to 100 Gbps per connection, adjustable in 10 Mbps steps |
| Services | E-Line, E-LAN, E-Tree, virtual cross-connect, cloud on-ramp, DIA, IP VPN |
| Cloud on-ramps | AWS, Azure, Google Cloud, Oracle Cloud; 50 Mbps to 10 Gbps hosted connections |
| Locations | 40+ PoPs; London, Frankfurt, Amsterdam, Ashburn, Mumbai, Singapore hubs |
| Change time | Under 5 minutes for bandwidth and new connections between on-net ports |
| APIs | REST/OpenAPI, Terraform, Ansible, webhooks; MEF LSO Sonata and Presto aligned |
| Telemetry | 1-minute metrics; gNMI/OpenConfig streaming; Prometheus, Splunk, syslog |
| SLA | 99.99% availability on protected paths; latency and jitter targets per route |
| Billing | Per Mbps-hour, committed-plus-burst or reserved; USD, EUR, GBP, INR |
How Connectivity is priced.
Circuits carry a monthly recurring charge set by bandwidth, access type and SLA tier; SD-WAN and managed LAN per site. Every site is quoted individually with a lead time.
We publish the model, not a public rate card - actual rates depend on destination, route class, volume and regulatory cost. See how every Dollu service is priced.
- MRC by bandwidth, access type and SLA tier
- Site-by-site quotes with lead times
- One-time install charge where applicable, quoted up front
- 12–36 month terms
- Multi-country consolidation discounts
- Billing
- Monthly recurring in advance; install charge on delivery; single invoice across countries
- Commitment
- 12–36 months per circuit
NaaS & On-Demand - your questions answered.
The questions customers and carriers ask us most often before they interconnect. If yours is not here, our team answers within one business day.
Still have a question?
Ask our solutions teamRelated services.
Services customers commonly combine with Network-as-a-Service (NaaS) & Network-on-Demand.
Provision your first connection this week.
Tell us where your data centres, clouds and sites are and we will map them to the fabric, quote a port and give you portal and API access to build the rest yourself.
