The ByteBlower server can be configured with multiple switches and switch types yielding a flexible testing solution. This article will look at the different intefaces and daisy chaining possibilites.

Network Interfaces

The ByteBlower GUI and API offer two types of network interfaces for traffic generation ("traffic interfaces"):

• Nontrunking interfaces are always found on the ByteBlower server. In general there are only few of this type but their bandwidth is not shared with any other interfaces.

• Trunking interfaces are located on the ByteBlower switches. Unlike the nontrunk, systems can have many of this type. They tend to share connectors on the ByteBlower server but have the benefit of increasing port density and offering more flexibility in connectivity (e.g. NBASE-T)

Since ByteBlower Server v2.14 individual trunking interfaces can be grouped together very flexibly. The examples below show some of the possibilities.  Much more is possible of course! This feature is still in development (December 2021), contact us through support.byteblower@excentis.com to convert your system to a new layout.

This article lists the new possibilities, for systems up to 2.13 we recommend reading this article.

Daisy chain configurations

Within ByteBlower, a daisy-chain is any setup where several switches are connected in series starting from a single ByteBlower server connector. The most common variant is where are a 1Gbit switch is extended with a smaller NBASE-T capable switch (setup pages).

For older ByteBlower versions, all daisy-chained switches were grouped together in a single large trunk. As a user you were required to perform the arithmetic where each interface was located.

In the new version this has become much easier, the interfaces of each switch can easily be placed together in the same trunk.

Multi-trunk

This setup is in many ways the reverse of the above: multiple connectors on the server chassis are connected to the same switch for larger total throughput.

Prior to v2.14, the above setup resulted in 4 different trunks, with the interfaces split evenly across them. Users thus needed to calculate themselves where each interface was located.

The new version makes it easier. All interfaces belonging to the same switch can just be grouped together in a single trunk.

More possibilities

The above two examples show just the 2 most common options. What else is possible?

• Pooling all DUTs of the same type in a single trunk. Regardless of where they are physically located

• Creating many small trunks. For example, one per modem.

• Splitting a shared ByteBlower System into different trunks, one for each team utilizing the setup.

The possibilities are limitless. Don’t hesitate to contact us if you don’t see your use-case mentioned. Together we’ll make it work!