With Pureport's Multicloud Fabric, you can easily connect your cloud environments, data center locations, and main/remote branches, forming a fully meshed, private, any-to-any network... without needing a dedicated BGP routing specialist!


Our self-service console and robust API allows your IT generalist, DevOps team, or network specialists to quickly build (and maintain) private connectivity. There are no long-term contracts and you don't have to order physical circuits or cross-connects.


Pureport locations and backbone

Pureport maintains points of presence in several key locations, world-wide. With Multicloud Fabric, our software-defined networking platform, you can connect each location to cloud services (such as AWS, Google, Azure, and others) or your own, local site (such as a data center or branch office). 


Additionally, you can leverage our redundant, high-speed backbone to enable low-latency communication among connections in different Pureport regions.


Customer networks

The basic construct in the Pureport model is the Network, which is the control plane. By creating a network, you can add Connections to your cloud environments and physical locations. You can build a single network to connect all of your environments, or segment into separate networks for different applications, workloads, organizational units, or user groups.


Connections within a single network are fully meshed at Layer 3 via the BGP routing protocol. As a result, when you deploy Connections in different locations (or even in the same location), the connected environments can communicate directly with one another. Essentially, the Network itself doesn't have a "location." All Connections added to a Network can communicate directly with one another, assuming communication is permitted by each environment's individual security settings. 


For example, if you add an AWS VPC, an Azure Vnet, and a physical location via IPSEC, each is able to send and receive traffic directly to any of the others. As shown in the following illustration, connections attached to all three Locations (Seattle, San Jose, and Ashburn) can communicate directly with each other:


By default, the Pureport Network allows all traffic. Use the security settings within your cloud environments and on-premises firewalls to allow the traffic or control specific traffic types.


Customer connections

After creating a network, you can add Connections for each environment to be able to communicate with each other. Pureport supports the following Connection types:

  • IPSEC tunnels for remote sites, using existing Internet connectivity
  • AWS Direct Connect
  • Azure ExpressRoute
  • Google Cloud Interconnect


Coming soon:

  • Physical ports, for either remote sites via 3rd-party private line (MPLS, point-to-point) or cross connect for customers who share data centers with our locations (coming soon)
  • Oracle FastConnect (coming soon)
  • Mobile VPN Gateway for remote workers (coming soon)


High availability

All Connections are Highly Available (HA) by default. They include redundant network paths through redundant equipment. For each new Connection added to your Network, two routing gateways are deployed across separate, redundant nodes on our platform and connected via redundant, high-capacity switches.


For workloads or applications that do not require High Availability, Pureport offers a limited selection of connection types and speeds without HA at a reduced cost. You may consider these for workloads where up-time is not essential, such as a development project. For most connections however, HA is mandatory and cannot be disabled in the console or via the API.


Regardless of your specific connections, the Network itself and Pureport's platform in general, are always HA to ensure maximum up-time. Pureport maintains redundant Internet connections, redundant backbone connections, and redundant private connections to the cloud providers (such as AWS Direct Connect) as well as high-capacity routing and switching engines. This enables you to build connectivity between sites and clouds and be assured of uptime and availability.