FAQ DirectCLOUD for Microsoft Azure

This article provides answers to frequently asked questions (FAQs) regarding the DirectCLOUD service for Microsoft Azure.

What are the differences between the Microsoft Azure Peering Service (MAPS) and ExpressRoute?

You need to take a close look at your own use case, but for a good initial assessment it is enough to take a understand the following points:

• ExpressRoute is a dedicated, private connection.
• Azure Peering Service is an optimized “public” connection type.

This means that private connections such as IaaS connections of virtual networks should be solved via ExpressRoute, while SaaS and PaaS applications that are optimized for access from public networks, such as Microsoft 365, Dynamics, or AzureDevOps, should preferably be connected via the Azure Peering Service.

Another difference is that, in contrast to ExpressRoute, no additional costs on the Microsoft side occur, neither for the service nor for the traffic or gateways involved.

What different billing models does Microsoft provide?

Microsoft provides three different options for handling bandwidth billing.

Metered model:

The most straight forward one is the metered model where you “pay as you go” per GB of egress traffic. The price ranges from 0,02 € to about 0,11 € depending on the Network Zone. You find an overview of all onramp locations and the matching zone here.

Unlimited model:

As the name suggests it includes the whole egress traffic here. Depending on which SKU you are using (Standard or Premium) also inside the whole Microsoft backbone. The standard SKU includes all traffic inside the geopolitical region, for example Europe, and the Premium SKU even includes worldwide traffic. But traffic from the Global Reach Add-on is always billed separately. Please calculate if it makes sense for you to use this option, because it can become very expensive.

You can do this with the Azure cost calculation tool.

Local Model:

The local model works almost the same as the unlimited option but with some slight differences which are important to know. The first difference is, it is only available for connections 1Gbps or higher. But the most important thing to know is that you can only connect vNETs from the same datacenter region. For example if you have the onramp-location Frankfurt it is not possible to connect networks from other regions than “Germany West Central”.

Is it possible to summarize the bandwidth of both VLAN's?

Microsoft is offering BGP Multi-path (multiple path selection enables BGP to load-balance traffic across multiple links).

Customer towards Microsoft: Both can be used!

Microsoft towards Customer: Both can be used!

One VLAN is for private peering and the other one for public peering, correct?

Yes, if you refer this to the 'inner VLAN'.

That means e.g. it always ends into a QinQ solution on the access. Outer VLAN 1 describes the Microsoft Link. The inner VLAN's describe the chosen service from Microsoft.

• DE-CIX Port A: VLAN 1.1 Public
• DE-CIX Port A: VLAN 1.2 Private
• DE-CIX Port B: VLAN 1.1 Public
• DE-CIX Port B: VLAN 1.2 Private

Are both VLAN's active/active or rather active/passive?

Both VLAN's are active. The control is taken place in BGP on Layer 3.

Is the second VLAN automatically taken over the control in case of outage?

Yes, totally correct. Control automatically on Layer3 in BGP.

Does both VLAN's have to be implemented on different Access Ports?

No, but it is highly recommended since its the only way to guarantee end to end redundancy.

Is it possible to just provision only one VLAN?

No, Microsoft's pre-requirement for all partners is to stipulate the provisioning of both VLAN's.

Need help? Get in touch with our customer service.