Technical Lab: Select an ExpressRoute Connectivity Model
Questions
Question 1 — Multiple Choice
A company has its own routers installed in a colocation datacenter shared with an Exchange provider. The network team wants to establish connectivity with Azure via ExpressRoute and has full control over BGP routing.
Which connectivity model is most suitable for this scenario?
A) Point-to-point Ethernet connection, as it offers a dedicated link between the datacenter and Azure.
B) CloudExchange colocation, as it allows the customer to establish virtual cross-connects with the Microsoft Enterprise Edge (MSEE) through an Exchange provider already present in the same datacenter.
C) Any-to-any (IPVPN), as it integrates offices and datacenters into an existing MPLS network.
D) ExpressRoute Direct, as it's the only model that allows BGP control in colocation environments.
Question 2 — Technical Scenario
A telecommunications company operates a global MPLS network that connects dozens of regional offices and two corporate datacenters. The architecture team wants to extend this connectivity to Azure without redesigning the existing network topology, treating Azure as another "site" in the corporate WAN network.
Office A ---+
Office B ---+----- MPLS Network (WAN provider) ----- ??? ----- Azure
Datacenter ---+
Which ExpressRoute connectivity model directly meets this requirement?
A) CloudExchange colocation, as it's already present in major datacenter hubs.
B) Point-to-point Ethernet connection, as it offers dedicated bandwidth between each site and Azure.
C) Any-to-any (IPVPN), as it allows the WAN provider to add Azure as a node in the existing MPLS network, preserving the hub-and-spoke topology already in use.
D) ExpressRoute Direct, as it offers 10 Gbps or 100 Gbps capabilities needed for large enterprise networks.
Question 3 — True or False
In the ExpressRoute Direct model, the customer establishes the physical connection directly to the Microsoft Enterprise Edge (MSEE) router ports at the peering facility, without intermediation from a layer 2 or layer 3 connectivity provider.
True or False?
Question 4 — Technical Scenario
An architect is comparing two connectivity models for a financial organization that requires multiple active ExpressRoute circuits simultaneously, needs 40 Gbps aggregate bandwidth, and wants to minimize the number of physical connections to manage.
Requirements:
- Bandwidth: 40 Gbps aggregate
- Simultaneous circuits: multiple
- Physical connections to manage: minimum possible
The architect considers using ExpressRoute Direct with 100 Gbps ports. What is the main technical advantage of this model compared to Point-to-point Ethernet for this specific case?
A) ExpressRoute Direct allows creating multiple ExpressRoute circuits over the same pair of physical ports, consolidating bandwidth without multiplying physical connections.
B) ExpressRoute Direct eliminates the need for BGP, simplifying routing for financial organizations with restrictive policies.
C) ExpressRoute Direct is the only model compatible with Microsoft 365 and public peering services, mandatory for financial organizations.
D) ExpressRoute Direct reduces latency compared to Point-to-point Ethernet because it uses dedicated fibers different from those used by other customers.
Question 5 — Multiple Choice
When selecting an ExpressRoute connectivity model, which of the statements below correctly describes a relevant technical difference between the Point-to-point Ethernet model and the CloudExchange colocation model?
A) Point-to-point Ethernet requires the customer to be physically in a colocation datacenter, while CloudExchange colocation can be contracted from any location.
B) In CloudExchange colocation, the layer 3 connection with the MSEE is always managed by the Exchange provider, while in Point-to-point Ethernet the customer can choose to manage BGP routing directly.
C) In Point-to-point Ethernet, the provider delivers a dedicated Ethernet link between the customer's facility and the Azure peering location, while in CloudExchange colocation the physical connectivity to the MSEE is intermediated by virtual cross-connects within the colocation facility.
D) Point-to-point Ethernet supports speeds above 10 Gbps, while CloudExchange colocation is limited to connections up to 1 Gbps due to shared infrastructure restrictions.
Answer Key and Explanations
Answer Key — Question 1
Answer: B
In the CloudExchange colocation model, the customer has equipment in a colocation datacenter where an Exchange provider also operates. Connectivity with Azure is established through virtual cross-connects between the customer's router and the MSEE. Since the customer controls their own equipment, they can manage BGP directly, without depending on a provider to route traffic.
The main misconception represented by the incorrect alternatives is confusing BGP routing control with physical connectivity model. Point-to-point Ethernet (C) can also support customer-managed BGP, but assumes the customer's facility is not in the same datacenter as the Exchange provider, meaning a dedicated physical link to the peering location is required. ExpressRoute Direct (D) is not exclusive to colocation, and its differentiator lies in direct connection to MSEE ports, not BGP control per se.
Answer Key — Question 2
Answer: C
The Any-to-any (IPVPN) model was specifically designed to integrate Azure with existing MPLS-based WAN networks. The WAN provider adds the Microsoft Enterprise Edge as a new location in the network, allowing all sites already connected to MPLS to reach Azure without topology changes. This preserves the corporate hub-and-spoke model and existing routing governance.
The conceptual error in the other distractors is confusing bandwidth or physical presence with topological integration. CloudExchange colocation (A) requires the customer to be in the same datacenter as the provider. Point-to-point Ethernet (B) would require a dedicated link per site, breaking the centralized topology. ExpressRoute Direct (D) is relevant for very high bandwidth requirements but doesn't solve integration with existing MPLS.
Answer Key — Question 3
Answer: True
In ExpressRoute Direct, the customer leases physical ports directly at Microsoft's peering facility (typically 10 Gbps or 100 Gbps per port, in pairs for redundancy). There is no layer 2 (Ethernet) or layer 3 (IP/MPLS) provider intermediating the physical connection. The customer is responsible for bringing their own physical connectivity to the peering facility and establishing BGP sessions directly with the MSEE.
This behavior is the opposite of other models (CloudExchange, Point-to-point, and Any-to-any), where an intermediate connectivity provider is responsible for the segment between the customer and the MSEE. ExpressRoute Direct is indicated for scenarios requiring full control of the physical layer, multiple circuits over the same ports, or regulatory requirements that prohibit third-party intermediation at the transport layer.
Answer Key — Question 4
Answer: A
The main technical advantage of ExpressRoute Direct in this context is the ability to create multiple logical ExpressRoute circuits over the same pair of physical ports. This means an organization with 100 Gbps ports can provision multiple 10 Gbps or 40 Gbps circuits sharing the same physical infrastructure, reducing the number of physical connections to manage without sacrificing aggregate bandwidth.
Point-to-point Ethernet requires a dedicated physical link per circuit, which to achieve 40 Gbps with multiple circuits would imply multiple physical connections managed separately.
The other distractors represent common misconceptions: BGP is not eliminated in ExpressRoute Direct (B); peering with Microsoft 365 is not exclusive to this model (C); and latency in ExpressRoute Direct is not reduced by a different fiber architecture, but by eliminating intermediaries in the physical layer (D), which is a secondary consequence and not the main differentiator for the described case.
Answer Key — Question 5
Answer: C
The correct technical difference lies in the nature of physical connectivity between the customer and the MSEE. In Point-to-point Ethernet, the provider provisions a dedicated Ethernet link that physically traverses the distance between the customer's facility and the Azure peering location. In CloudExchange colocation, the customer is already physically in the same datacenter as the Exchange provider, and connectivity is made through virtual cross-connects within that facility, without needing a long-distance circuit.
Alternative B incorrectly describes BGP management: in both models, layer 3 can be managed by the customer or provider, depending on the service agreement. Alternative A reverses the physical location premises. Alternative D is factually incorrect: CloudExchange colocation supports speeds up to 10 Gbps, and the 1 Gbps limitation is not an inherent restriction of the model.