Network Fabric
MEF-compliant DIA, MPLS/IPLC/EVPL private lines, SD-WAN overlay with CoS, multi-cloud direct connects, and bonded last-mile failover — the transport layer for every Clevertek solution.
What Network Fabric is
The Network Fabric is the physical and logical transport layer that every other Clevertek solution builds upon — dedicated internet circuits, private MPLS links, SD-WAN overlays, multi-cloud direct connects, and managed Wi-Fi LAN, delivered and operated as a single programmable fabric. Traditional multi-carrier WAN architectures introduce failure at the seams: a single carrier routing incident becomes a branch-site outage, a cloud direct-connect link saturates during a migration window without dynamic path adjustment. We engineer the fabric as carrier-neutral by design, with dual-carrier failover paths, BFD-enabled sub-second convergence, and bonded last-mile aggregation ensuring that a cable cut, carrier BGP withdrawal, or site-power event never reaches the user's application session. The carrier relationships, SLA monitoring, and fault escalation are all managed within the fabric operating model — you make one operational call rather than chasing three carriers about whose segment caused the degradation.
How Network Fabric is composed
Every Clevertek solution rides on the Network Fabric, so its behaviour sets the ceiling for everything above it. These are the layers it is built from and the way it is operated once live.
Access and last mile
Dedicated internet access, leased lines and bonded circuits, sized per site against the traffic profile that site actually carries. Where the building allows it we procure physically diverse routes from separate carriers, because two circuits sharing one duct are one circuit with extra paperwork.
Private transport
MPLS, IPLC and EVPL links for site-to-site and India-to-global traffic that needs committed bandwidth and class-of-service treatment. CoS marking keeps voice and interactive traffic in a protected queue while bulk replication absorbs what is left.
Internet and carrier peering
IP transit delivered over carrier-neutral interconnect, with carrier aggregation across multiple upstream providers. Prefix advertisement and path selection are managed by us, so a single upstream withdrawing a route does not become a customer-visible event.
Overlay and policy
An SD-WAN overlay that measures latency, jitter and loss per path and steers each application class onto the transport that meets its own budget. Policy is defined once at the controller and applied consistently to every site.
Cloud and data-centre interconnect
Direct connects into public cloud regions, colocation facilities and partner data centres, so traffic between your estate and a hyperscaler does not traverse the public internet by default.
Campus and branch LAN
Switching, managed Wi-Fi and edge routing supplied and patched under the same contract as the WAN, with segmentation between guest, corporate and operational device classes.
How we run it
Design is the smaller half of the work. These are the operating practices that keep Network Fabric behaving as designed once it is live.
Monitoring against the committed SLA
Each path is probed continuously for latency, jitter and packet loss and compared against the committed figure. The alarm fires on the degradation trend rather than after users report an outage.
Carrier escalation as our job
We hold the carrier relationships. When a circuit degrades we raise the ticket, chase the provider and report progress, so your team is not relaying messages between three vendors who each believe the fault sits elsewhere.
Change control across the estate
Routing changes, policy changes and circuit additions go through one review. A change on one fabric is checked against the security and cloud fabrics before it ships, which is what stops a routing fix from quietly opening a path.
Capacity and lifecycle review
Utilisation against forecast is reviewed on a cycle, so circuit upgrades and equipment refreshes are planned purchases rather than emergency procurement after a saturation incident.
Products on Network Fabric
Each product is available as a managed service on this fabric.
Frequently asked questions
Do we have to replace our existing circuits to use the Network Fabric?
No. The fabric is layered over what you already run. Existing MPLS or leased-line contracts are onboarded where they still meet the requirement, and new access is added where they do not. Because the fabric is carrier-neutral, the decision about which carrier supplies a site is made on coverage and cost rather than on what is already installed.
How does the fabric behave when a carrier fails?
Where a site has a second path, failover is automated through BGP with the overlay steering traffic onto the surviving transport. The hybrid paths stay active rather than sitting idle as cold backup, so failover is a traffic shift rather than a cold start. Where a site currently has only one path, we will quote a second and tell you what it changes.
Which parts of the fabric can we manage ourselves?
The split is agreed at design time. Some customers take the circuits and the monitoring and keep their own overlay policy; others hand us the full estate including LAN and Wi-Fi. Whichever split you choose, the monitoring, escalation and reporting stay with us, because fabric-level visibility is what makes the fault tolerance measurable.
Can the fabric extend to a new site or a new cloud region quickly?
A new branch or cloud region is a configuration change on the overlay plus whatever access circuit that location needs. The overlay change is ours and moves at our pace; the access circuit moves at the carrier timeline for that building, and we will tell you which of the two is the constraint before you commit to a date.
Architect on the right fabric
Tell us your workload — we will map the right products across the fabric.