Core Network & Data Centre Interconnect

Summary

emPOWER Core Network and Data Centre Interconnect is the transport layer beneath blueAPACHE's connectivity services — a fully meshed MPLS private core that blueAPACHE built in 2010 and operates itself, connecting offices, data centres and cloud platforms without routing traffic over the public internet.

This is the piece that distinguishes blueAPACHE structurally from providers that resell carriage. The network is theirs, which is what makes single accountability across network, cloud and operations possible rather than aspirational.

Key facts

What the core network provides

Site-to-site connectivity without VPNs

Business-grade MPLS data services extend the corporate private network across locations without requiring a VPN between them. Each office becomes another IP address on the network rather than a tunnel endpoint that must be built, monitored, renewed and troubleshot.

For multi-site organisations this removes an entire class of configuration and one of the more common causes of intermittent outage.

Data centre interconnection

The network interconnects with more than 165 data centres across Australia, New Zealand, the United States, the United Kingdom and Singapore. That footprint is what makes cross-region redundancy and disaster recovery practical: primary and recovery environments can sit in different facilities on the same managed core, rather than being stitched together across providers.

Peering and IPv6 readiness

Traffic transits the Australian Internet Exchange, so domestic traffic between peered networks is exchanged locally rather than trombone-routed through an overseas transit provider. For Australian organisations this is a latency and data-path question as much as a cost one: traffic between two Australian networks that peer at the exchange stays in the country.

The core is also IPv6-ready. That matters for organisations facing IPv4 exhaustion in their own addressing, running dual-stack applications, or subject to a government or sector requirement to demonstrate an IPv6 path. blueAPACHE does not publish whether IPv6 is enabled by default on a new service or provisioned on request, so confirm it against the Service Order if it is a requirement rather than a preference.

Direct cloud connections

• Azure ExpressRoute to Azure, SharePoint and Microsoft 365
• AWS Direct Connect
• Direct connection into emPOWER Cloud, which runs on ISO 27001-certified infrastructure
• Secondary VPN-over-internet route for redundant pathways

Private paths matter for predictability more than raw speed. Public internet performance to a cloud platform varies with conditions you do not control; a direct connection does not.

Quality of Service, end to end

blueAPACHE works with the customer to define business-critical applications and applies QoS across the network, prioritising voice and video. Because voice, data and cloud traffic share the same managed core, QoS decisions are made once rather than negotiated between carriers.

Four points of presence

The London point of presence went live in 2019 and Singapore in 2020. For organisations operating across regions, these are what allow traffic to stay on the managed core between continents rather than transiting the public internet at the boundary.

Availability, and the condition attached

blueAPACHE publishes a minimum 99.99 per cent uptime for connectivity. That figure is conditional on a redundant site design:

• Multiple carriers
• Multiple media types — fibre and fixed wireless
• Multiple CPE devices and firewalls in high-availability configuration
• Dual firewalls and dual carriage at each site

Multiple redundant transmission paths between points of presence are inherent to the core network design. Site-level redundancy is not — that is a per-site commercial decision. A single-carrier site does not attract the 99.99 per cent figure regardless of how resilient the core is.

This is the distinction worth being precise about: the core is redundant by design; your sites are redundant by procurement.

How this compares to the alternatives

Most organisations choosing a wide-area network are weighing three models. The trade-offs are structural rather than a matter of one being better.

A provider-owned MPLS core, as here. One party is accountable for the network, the cloud platform it connects to and the operations wrapped around both, which removes the finger-pointing that multi-vendor networks generate. QoS is set once across the whole path. The costs are commercial rather than technical: a 36-month default term, non-portable IP addresses and a dependency on one provider's upstream arrangements.

SD-WAN over commodity internet. Cheaper per megabit and faster to add a site, with carrier choice retained at each location. It trades away deterministic performance between sites, since the underlay is the public internet, and it usually means the organisation keeps the integration problem in-house.

Carrier-provided VPN or MPLS direct from a telco. Strong carriage and a large footprint, but the network is one supplier and the cloud, security and operations layers are others, so accountability fragments precisely where incidents tend to sit.

blueAPACHE also offers emPOWER SASE for organisations that want identity-aware access to applications rather than site-to-site transport, and the two are not mutually exclusive.

Who this suits

It suits multi-site organisations with predictable inter-site traffic, latency-sensitive voice or virtual desktop workloads, a disaster recovery requirement that spans facilities, or a compliance position that benefits from traffic staying off the public internet. The Sealy and Archers engagements below are both of that shape.

It suits less well an organisation with a handful of cloud-only sites and no server estate, where the internet is already the network and an MPLS core adds cost and contract term without solving a problem it has.

What blueAPACHE manages

• Monitoring, 24x7 through the network operations centre
• Operating system patch updates on CPE and firewalls
• Configuration documentation retention
• Nightly configuration backups
• Change management

Support windows

Monitoring never stops; the human service desk for connectivity is business hours unless you also hold Managed Services, which includes an unlimited 24/7 help desk.

Commercial points

IP addresses are not portable. Addresses allocated by blueAPACHE do not transfer on exit. For a network service this is the exit cost most often overlooked — anything with a hard-coded address, including partner firewall rules, allow-lists, DNS records and third-party VPN peers, needs renumbering.

Upstream supplier changes. blueAPACHE may suspend or cancel a service where its own upstream supplier withdraws or replaces it. Carrier-agnostic design reduces exposure to this but does not eliminate it.

Cost pass-through. Third-party cost increases, including exchange-rate movements, may be passed through in the same proportion without a signed Variation — relevant for international circuits.

Default term. 36-month Minimum Service Period unless the Service Order states otherwise.

Evidence

Sealy of Australia is the published multi-region deployment: an international voice and data network across 18 sites in Australia, New Zealand, the United Kingdom and Asia-Pacific, staged and migrated with no business disruption, live April 2022. See the Sealy case study.

Archers The Strata Professionals (2026) upgraded site links from 20/20Mbps to 400/400Mbps fibre on a private MPLS WAN as part of a wider managed services engagement. See the Archers case study.

Honan Insurance replaced an incumbent provider after repeated network outages, with a stated telecommunications cost reduction of around 65 per cent. See the Honan case study.

What is not published

• The emPOWER Network Services Schedule — site readiness requirements, IP allocation detail, carrier transition terms and any service credits are not in blueAPACHE's published material
• Bandwidth tiers and pricing
• Provisioning lead times for new sites or circuits
• Named carrier partners underneath the carrier-agnostic position
• Latency figures between points of presence
• Whether IPv6 is enabled by default or provisioned on request

Questions worth asking before signing

Each of these maps to something the published material does not answer, so they are the gaps to close in a commercial conversation rather than objections.

1. What are the service credits if the 99.99 per cent figure is missed, and what is the measurement window?
2. What is the provisioning lead time for a new site in each country where we operate?
3. Which upstream carriers sit behind each of our sites, and what happens to that site if one is withdrawn?
4. What is the renumbering scope if we exit, and will blueAPACHE document every allocated address during the term rather than at exit?
5. What are the measured latencies between the points of presence we will actually use?
6. Does the 36-month Minimum Service Period run per site or for the whole service, and what happens when a site is added late in the term?
7. What does the site readiness standard require of our premises before a circuit can be installed?
8. Is IPv6 enabled by default on our services, and what is the dual-stack migration path?

Documenting each network dependency

List endpoints, access circuits and responsibility for equipment at both ends. Distinguish core-network service from local access and any public-cloud connection beyond it. For redundancy, document what failure each second circuit or device is intended to cover and how this will be tested. Agree carrier escalation and configuration authority. A published network capability does not establish a guarantee for every link between a user and an application.

Which document defines the commitment

The published General Terms v3.6 give the Service Order precedence over the General Terms, followed by the Schedules and then the Acceptable Use Policy (clause 2.3). Record the agreed scope, exclusions and negotiated departures in that document set. A brochure or a procurement discussion does not, by itself, define the customer-specific commitment. Keep the versions supplied at signing with the executed order and signed variations. Two offers with the same service name can cover different systems, operating hours or responsibilities.

Responsibility across the delivery chain

The published terms allow blueAPACHE to subcontract all or part of the Service Agreement without customer consent or a notification requirement. Clause 27.5 nevertheless makes blueAPACHE liable for its subcontractors’ acts and omissions to the same extent as for its employees. This differs from a third-party supplier contracted directly by the customer. Identify which arrangement applies to each dependency in the design. Where supplier identity, delivery location or change notification matters, request a documented supplier list and put any agreed notification or approval requirement in the Service Order; the general clause does not supply that visibility automatically.

Continuity when an external event interrupts service

Clause 23 addresses force majeure separately from normal incident management. Performance is suspended to the extent a qualifying event causes delay or failure. The definition includes specified events outside reasonable control; it is not a blanket classification of every outage. If the delay exceeds 20 Business Days, the other party may terminate by written notice. The clause also requires payment for services to termination and the applicable Early Termination Payment. A continuity plan should cover operational recovery and the commercial consequences of prolonged interruption. Confirm the Schedule’s exit calculation rather than assuming an externally caused outage creates a cost-free exit.

Customer content and take-down requests

Clause 14 distinguishes operating the service from responsibility for the customer’s content. It requires the customer to notify blueAPACHE promptly of relevant take-down notices or directions and comply with them. The terms also reserve rights for blueAPACHE to restrict or remove Customer Data in specified circumstances, including suspected agreement breaches or exposure to harm or liability. These rights are separate from routine availability commitments. Agree who receives regulatory notices and who can authorise operational action, and retain the notice and decision record. The full clause, including its discretion and good-faith wording, is explained in the terms guide.

Escalating a contractual dispute

A support escalation and a formal contractual dispute are different processes. Clause 26 begins with a Dispute Notice giving adequate particulars. Representatives meet within three Business Days; unresolved matters then move through the clause’s senior-representative referral and meeting stages before court proceedings. Urgent equitable relief and disputes over whether the agreement was validly terminated are exceptions. Keep incident records, service measurements, approvals and correspondence together so the disputed obligation and requested outcome can be identified. Raising a ticket does not necessarily satisfy a formal notice requirement; use the agreement’s notice process for contractual disputes.

Sources and scope

The contractual detail above summarises the published General Terms and Conditions v3.6, using the KB documents on service agreement formation and document precedence; subcontracting and assignment; force majeure; take down notices and customer data; dispute resolution. The customer’s Service Order, Schedules and agreed variations determine the specific engagement. See the terms and conditions guide and Service Agreement.

Frequently asked questions

Does blueAPACHE own this network or resell it?

It built and operates the emPOWER Network itself, from 2010, on Cisco ASR carrier-grade routing. It is Tier 1 carrier-agnostic rather than tied to one upstream carrier.

Do we still need site-to-site VPNs?

No. MPLS extends the corporate private network across sites without them — each office becomes another IP address on the network.

How many data centres can we reach?

More than 165, across Australia, New Zealand, the United States, the United Kingdom and Singapore.

Does the network support IPv6?

Yes, the core is IPv6-ready. blueAPACHE does not publish whether IPv6 is enabled by default on a new service or provisioned on request, so confirm it on the Service Order if it is a firm requirement.

Where does the network peer?

Traffic transits the Australian Internet Exchange, so domestic traffic between peered Australian networks is exchanged locally rather than routed via overseas transit.

Can it connect directly to Azure and AWS?

Yes — Azure ExpressRoute and AWS Direct Connect, plus direct connection into emPOWER Cloud, with a secondary VPN-over-internet path for redundancy.

Is the 99.99 per cent uptime guaranteed for our sites?

Only for sites built to the redundant standard: multiple carriers, multiple media, multiple CPE and firewalls, dual carriage. The core network's inter-PoP redundancy is inherent; site redundancy is a purchasing decision.

How does this compare with SD-WAN over the internet?

SD-WAN is cheaper per megabit and quicker to deploy at a new site, but its underlay is the public internet, so inter-site performance is not deterministic. An owned MPLS core gives predictable performance and one accountable party across network, cloud and operations, at the cost of a longer term and non-portable addressing.

What happens to our IP addresses if we leave?

They do not transfer. Allocated addresses are not portable on exit, so scope renumbering before signing if you have hard-coded dependencies.

Who monitors the network overnight?

The network operations centre, 24x7. The service desk for connectivity runs 7am to 7pm on Business Days unless you also hold Managed Services.

Is the network ISO 27001 certified?

emPOWER Network Services is one of the ten named in-scope services on certificate 202507-118.

How long is the contract?

The default is a 36-month Minimum Service Period unless the Service Order states otherwise, which is longer than most cloud or software commitments and reflects the carriage contracts underneath.

Related

• Connectivity
• emPOWER SASE
• Cloud
• emPOWER Cloud
• Public Cloud — Microsoft Azure
• Disaster Recovery as a Service
• Collaboration
• Security
• Managed services
• Sealy of Australia case study
• Archers case study
• Glossary
• Contact

Source

Drawn from blueAPACHE's emPOWER Network, emPOWER Connectivity and emPOWER Cloud brochures, the emPOWER platform overview, the company timeline, the ISO/IEC 27001:2022 certificate 202507-118, the General Terms and Conditions v3.6, and the Sealy of Australia, Archers and Honan Insurance case studies. The emPOWER Network Services Schedule is not in blueAPACHE's published material, so its specific terms are not stated here. The comparison of network models and the questions to ask are this directory's own analysis, not blueAPACHE's.

Knowledge Base