NE-ONE: Functional Questions (Enterprise)

1-) What is a Soft Port and why are they needed?

There are numerous reasons why depending on the project requirements:

Soft Ports are a great way of sharing physical ports when many ports are needed but the I/O through each port is relatively moderate – in this way they are like Cisco’s sub-interfaces. This is useful in many marketplaces.
Layer 2: Soft ports can map to vlans allowing the creation of a test network in an environment where IP addresses cannot be modified and are often in the same subnet. In this case each system is put into a different vlan and only the NE-ONE allows the system to communicate with each other. This is of particular use in Defence.
Layer 3: IP (IPv4) soft ports can be set up act as gateways, so that large networks with independent subnets can be created without requiring external routers. This is of particular use in Defence, Home environments (including Gaming and IoT), SD-WAN testing/PoCs and many others. VLAN and IP can be combined for more power.
Filter Soft Ports allow arbitrary division of Ports (HW and Soft Ports) to allow a port, for example an Internet gateway, to be shared by multiple users.

2-) How do Network Objects affect licensing?

Each node and link on the diagram count as a network object. Each row in a cloud table (Cloud and the new TDMA cloud) count as 0.5 objects. Network object licencing allows the NE-ONE to be sold relatively inexpensively for simple tasks but as the complexity grows the licence price reflects the complexity of the task.

3-) What’s intelligent about Intelligent Packet Replay?

We have two forms of packet replay, one of them PPR (Passive Packet Replay)just plays back packets according to their timing in the pcap file. IPR (Intelligent Packet Replay) implements a TCP stack and maintains causality so that transactions that should follow each other continue to do so even in the event of bandwidth limitation, latency and loss. This means that as a background load the replay will behave correctly, as a real background load would do and as a foreground test, measurement of the times taken for the replay are a good indicator of what would happen if the application were run in the changed (bandwidth restricted, latency, lossy) environments.

4-) What’s the difference between a full mesh network and the Cloud?

A full mesh has a complex diagram potentially with every connection showing; after about six nodes the full mesh diagram can be very busy indeed.

The Cloud feature allows the user to draw a network as a network engineer would – with a blob in the middle (the cloud) representing a core network, for example the Internet, MPLS, etc. Inside the Cloud links are defined to represent the internal behaviour of the core network. This allows a fully or partially meshed network to be drawn as though it was a hub and spoke greatly reducing complexity.

Advantages of NE-ONE's Cloud feature

Simplicity
Drawings that match network engineers view of the world.

Disadvantages of the Cloud

In a mesh network you can put impairments on any link, whereas the cloud only supports bandwidth control, queue size, latency, jitter (random) and loss.

TDMA Cloud allows a core TDMA, timeslot-based core to be defined – particularly useful in defence ad-hoc networks

5-) Can more than one person use the NE-ONE for different testing?

Yes - multiple users can use NE-ONE each running their own separate test network.

Each user needs at least two ports (HW/vNIC or Soft Ports), and this is where soft ports also come in very handy.

6-) What are Virtual Routers?

Nodes (picture objects) all can route using several routing functions, they can therefore be thought of as Virtual Routers. However, they do not emulate specific router (e.g. Cisco) models and today use static routing, though dynamic routing is on the roadmap.

7-) What’s the minimum amount of bandwidth that you can configure for a link?

1bps

8-) What is the maximum amount of latency that you can configure for a link?

10 Gbps (for current models though in practice there are likely to be packet rate limitations depending on the configuration)

9-) Is it possible to configure different impairments for the uplink compared with the downlink?

Yes, the directions are independent.

10-) What GUI functions does the RESTful API Support?

All of them. There is no separate User RESTful API. Users get access to the full RESTful API that the GUI uses and in fact can ‘learn’ the calls from the GUI by using the web browser’s inspector feature.

11-) Can I add my own impairments?

Yes, you can. However, this involves learning the language that impairments are written in (PEPL).

12-) Is an Internet connection required to operate the NE-ONE?

While the NE-ONE is web browser driven, it does not need to be connected to the Internet for any purpose. Documentation, Help videos, etc are all on board. Updates to the NE-ONE software and underlying operating system all come through the software update mechanism which does not require Internet access. Licence keys also do not require the Internet for verification.

13-) Is the Virtual version the same as the hardware version?

There is one code base for all versions. The product automatically detects whether it is running on Hardware, a Virtual Machine or in the Cloud (AWS, Azure) and configures itself for that environment. So, once you have learned how to use NE-ONE in one environment then it will operate the same way in another.

Having said that the environments do create some restrictions, for example:

VMware switches do not behave like hardware switches; ports do not ‘learn’ what is connected to them (i.e. their CAM table is statically learned from what is connected in VMware). This means they must be configured with specific port/port group settings to work correctly.
The cloud platforms (AWS, Azure) do not present virtual switches at all, rather subnets – think in terms of network segments with a built-in core router. VLANS are not available in these environments and routing must be configured to use NE-ONE as the router, so that impairments can be created between the subnets.