...

If you are attending a VSPERF-training session (ex: IEEE NFVSDN tutorial), kindly download and install following, before the training-sessions. 

(Note: Of the three options described below, we'll only use Option-1 for IEEE NFVSDN Tutorial).

docker pull opnfv/nfvbench

The Installation and configuration details are provided below.

Option-1: Experimenting VSPERF with Oracle Virtualbox:

You would need two VMs in this scenario. Figure below shows the setup.

Step-1: Downloading VM-Images

Download following images for the two VMs:

1. DUT-VM: Any Linux image. Ex: Ubuntu 16.04 (http://releases.ubuntu.com/16.04/ubuntu-16.04.5-desktop-amd64.iso) 
2. TGen-VM: http://trex-tgn.cisco.com/trex/T_Rex_162_VM_Fedora_21.ova. 

Step-2: Installing VMs:

1. DUT-VM: This is just a standard VM-from-ISO installation procedure. You will find many step-by-step description (Ex: ,  https://itsfoss.com/install-linux-in-virtualbox/) of this, including Videos (Ex: https://www.youtube.com/watch?v=QbmRXJJKsvs) . 
   1. Please ensure that number of interfaces are 4 and configured according to the figure above.
2. TGen-VM: The details of installation of the OVA can be found here: https://trex-tgn.cisco.com/trex/doc/trex_vm_manual.html, under TRex inside Virtual Box.
   1. Please ensure that number of interfaces are 4 and configured according to the figure above.

The names of the interfaces maybe different - for example on Ubuntu it may be enp0s2/3/4...

Step-3: Copying SSH keys between VMs.

To ensure smoother access, ensure DUT-VM can ssh to TGen-VM without password-probe, copy the SSH public keys. The step to follow:

1. If no identity is created yet (ex: if you don't find id_rsa.pub in ~/.ssh folder) Run ssh-keygen. It will generate the necessary identify.
2. Use ssh-copy-id command. ssh-copy-id username_of_tgen_vm@ipaddress_of_tgen_vm
   
   

Step-4: Installing VSPERF on DUT-VM

Instructions to install and configure (including tuning) VSPERF can be found here 

...

1. mkdir -p /mnt/huge
2. mount -t hugetlbfs nodev /mnt/huge
3. echo "vm.nr_hugepages=64" >> /etc/sysctl.conf

Step-5: Updating T-Rex on TGen-VM

Download the latest T-Rex: wget --no-cache http://trex-tgn.cisco.com/trex/release/latest. to the home (/home/trex) folder.

Untar it.

Step-6: Docker Installation on DUT-VM

Install Docker on DUT-VM. You can follow the steps described here for Ubuntu. 

Step-7: Running Tests

Instructions to run tests can be found here.  This would be covered during the training session.

Option-2: Experimenting with Single Physical System 

In this scenario, both the DUT (vswitch with/without VNFs) and the Traffic-Generator run on same System.

1. vsperf distribution: Downloaded from here
2. OS: Linux (ex: centos, fedora, suse, ubuntu, RHEL)
3. Hypervisor: Qemu (> 2.3)
4. RAM:  8-16GB
5. TGen-VM: TRex-VM (VSPERF includes this image as part of its artifacts - https://artifacts.opnfv.org/vswitchperf/vnf/vloop-vnf-ubuntu-16.04_trex_20180209.qcow2)
6. VNF: vloop-VNF (VSPERF includes this image as part of its artifacts - http://artifacts.opnfv.org/vswitchperf/vnf/vloop-vnf-ubuntu-14.04_20160823.qcow2)

Option-3: Experimenting with Two Physical Systems

In this scenario, DUT runs on first system and the Traffic-generator runs on a different system. (Note: The second system could be a hardware traffic-generator - Spirent, Ixia, Xena, etc.)

System-1:

1. vsperf distribution: Downloaded from here
2. OS: Linux (ex: centos, fedora, suse, ubuntu, RHEL)
3. Hypervisor: Qemu (> 2.3)
4. RAM:  8-16GB
5. NICs: 2 for data-traffic
6. VNF: vloop-VNF (VSPERF includes this image as part of its artifacts - http://artifacts.opnfv.org/vswitchperf/vnf/vloop-vnf-ubuntu-14.04_20160823.qcow2)

System-2:

There are multiple options here - user can choose any one of the following

...