Riverbed virtual products are available for download from the Support site. Access to download packages requires that you have a registered user account. Downloadable deployment packages are archived in formats suitable for a variety of virtualization environments. Each package contains:

The steps to deploy any virtual appliance are generally the same across products:

As you provision resources to the virtual machine, provision the amount required for the model plus an additional amount for hypervisor overhead. For ESXi, reserve the memory and CPU cycles needed for the appliance model and verify that the host has resources to accommodate the 5 percent hypervisor overhead. For Hyper‑V, reserve the memory and CPU percentage needed for the appliance model and verify that the host has 1.5 GB and 15 percent CPU remaining, for hypervisor overhead.

The deployment package for VMware ESXi is an OVA archive, and it contains the VMX and VMDK files that are necessary to create the virtual appliance. The VMX file is the appliance specification file. The VMDK files represent the appliance’s management and data store virtual disks. The package contents require several gigabytes of disk space. Do not modify any of the files in the package.

The ESXi datastore, as distinct from the appliance’s data store, is part of the underlying virtual environment and provides the virtual storage where the appliance’s management and data store disks are to be located. Make sure that the datastore has enough capacity for the OVA package. You can install the smaller VMDK containing the management disk on an ESXi datastore backed by any type of storage media. We recommend that you put the larger VMDK containing the appliance’s data store on a datastore backed by the fastest available physical storage media. That datastore should have enough room to expand to the size required by the appliance model.

If you’re using Riverbed network interface cards (NICs) on the physical ESXi host, you must map the appliance’s LAN interface to the virtual machine’s pg-vmnic2 port label and the appliance’s WAN interface to the virtual machine’s pg-vmnic3 port label.

The deployment package contains predefined configuration for the default model appliance, which is model VCX30. You’ll need to first install the default model, and then upgrade to your desired VCX model.

The following table shows supported RiOS versions for each listed ESXi version.

Some models only support a two-port 10-GbE multimode fiber NIC using direct I/O. You can configure bypass support using the VMware Direct Path feature. This feature allows the appliance to directly control the physical bypass card. You must use a Riverbed NIC. If you currently use a Riverbed NIC with ESXi, you can use the same card if you want to upgrade the ESXi version.

The deployment package for Microsoft Hyper‑V is a ZIP archive. You’ll need to run the installation script in the deployment package from Windows PowerShell. To run the script, you might need to configure the virtual machine’s security policy to Unrestricted using the PowerShell Set-ExecutionPolicy Unrestricted command. This table lists the parameters for the installation script.

Before you begin, create and connect the necessary virtual interfaces and switches.

We recommend that you enable MAC address spoofing on the virtual network adapters. You’ll need to run PowerShell as an administrator.

where <vm-name> is the name of the virtual machine, <hyper‑v-host> is the name of the Hyper‑V host, and <v-network-adapter> is the name of the virtual network adapter.

This example shows how to enable MAC address spoofing for the lan0_0 and wan0_0 interfaces on the virtual machine myVM and Hyper‑V host myHost:

Verify that MAC address spoofing has been enabled:

where <vm-name> is the name of the virtual machine and <hyper‑v-host> is the name of the Hyper‑V host.

This example shows how to display MAC address spoofing status for the configuration as a formatted list:

Determine if your network uses VLAN tagging before you deploy the appliance. Specific configuration is required depending on the VLAN configuration. This table shows the configuration to use for networks with and without VLAN tagging.

After powering on the virtual machine, if you see messages about missing interfaces or disks, check these troubleshooting tips:

After every change to Hyper-V configuration settings, you must shut down and restart the Hyper-V.

Be sure that the lan0_0 and wan0_0 interfaces are mapped to the correct NIC on the server.

The deployment package for the KVM package is a TAR archive. Kernel-based Virtual Machine (KVM) is a virtualization solution for Linux on x86 hardware. A KVM consists of a loadable kernel module that provides the core virtualization infrastructure and a processor-specific module that provides virtualization extensions. Using KVM, you can run multiple virtual machines running unmodified Linux or Windows images. KVM is open source software. The kernel component of KVM is included in mainline Linux, as of version 2.6.20. The user-space component of KVM is included in mainline QEMU, as of version 1.3.

KVM supports various I/O virtualization technologies. Paravirtualized drivers, which enable direct communication between hypervisor-level drivers and guest-level drivers, provide the best performance when compared with full virtualization. The virtio API provides a common set of paravirtualized device drivers for KVM.

Virtual appliances for KVM can be deployed in different ways, each method using a different procedure. This document describes how to deploy appliances for KVM by using the installation script supplied in the product’s deployment package and the virsh command.

Ensure the KVM host system has at least four network interfaces, and that the system’s network environment is configured so that the LAN and WAN interfaces are on separate networks or bridges.

Resizing the data store might be required if your virtual SCC is managing more than ten appliances.

Because the virsh reboot and virsh shutdown commands are not support by virtual SCC, you’ll need to use the virsh destroy command. Run these commands to resize the data store disk:

To deploy on Linux, run the installation script, specifying values for the:

Create the virtual appliance by entering the virsh create <virtual-appliance-name>.xml command.

Start the appliance by running the virsh define <virtual-appliance-name>.xml command followed by the virsh start <virtual-appliance-name>.xml command.

The deployment package for the Cisco Enterprise Network Compute System (ENCS) is the same TAR archive used for Linux. ENCS is a line of compute appliances designed for the Cisco SD-Branch and Enterprise Network Functions Virtualization (ENFV) solution. Cisco SD-Branch is a hosting platform designed for the enterprise branch edge. The platform provides a virtual environment that enables the automated deployment of virtual network services consisting of multiple virtualized network functions (VNFs). Using the platform, administrators can leverage the flexibility of software-defined networking (SDN) capabilities to service chain VNFs in a variety of ways. Cisco SD-Branch is comprised of these components:

In this context, Riverbed products serve as VNFs running as virtual appliance on NFVIS. In-path deployments and out-of-path deployments using Web Cache Communication Protocol (WCCP) and policy-based routing (PBR) are supported. While appliances can be deployed in different locations in a topology depending on your needs, the procedures here focus on an in-path deployment. After you understand the underlaying concepts, you will be able to design and execute different kinds of deployments.

Before deployment, ensure that the:

Download the package to a system running any supported Linux operating system. You will prepare the software image for use on Cisco 5000 Series ENCS on there.

You can use the Riverbed-provided helper script—together with Cisco helper files—to create an image file that can be immediately uploaded to NFVIS and deployed. This method provides some flexibility in setting the attributes of the virtual machine and automates much of the process. You can also manually prepare the image. This method requires more steps but provides the most flexibility in configuring virtual machine attributes.

After you add an image to the NFVIS image repository and register it, you can use the image on any Cisco NFVIS system.

The script helps to automate the image preparation and packaging process while allowing you some flexibility in setting virtual machine properties. The script generates a .tar.gz image file to your specifications that you can upload and deploy.

The Riverbed helper script requires two additional files that you obtain from Cisco. You must place these files in the same location as the Riverbed helper script:

Unzip and untar the product’s deployment package, and then log in to your account on NFVIS. Choose VM Life Cycle > Image Repository > Browse Datastore > Data > intdatastore > Uploads > vmpackagingutility > nfvisvmpackagingtool.tar.

Download the nfvisvmpackagingtool.tar file to your local system. Unpack the nfvisvmpackagingtool.tar file, and locate these files: image_properties_template.xml and nfvpt.py. Place these two files in the same location where you placed the Riverbed helper script, riverbed_encs_package_gen.py.

Run the Riverbed helper script and follow the prompts. The system creates a .tar.gz file that is suitable for upload to the NFVIS image repository.

In the NFVIS console, choose VM Life Cycle > Image Repository: Image Registration. Upload the .tar.gz file to the repository and then register it.

Follow this procedure if you want more flexibility in setting virtual machine attributes. You’ll need to:

After you have prepared these files, you can import them into NFVIS using the Image Packaging section of the NFVIS console. There you can package the uploaded files into an image (.tar.gz file) suitable for use on any Cisco 5000 Series ENCS. After packaging, register the image in the repository.

To upload images to the NFVIS image repository, log in to your account on NFVIS. Choose VM Life Cycle > Image Repository, and select the Image Packaging tab. Click the icon next to VM Packages, and then enter values for these fields:

Package Name is the name for this instance of the SteelHead package.

VM Version is the version for this instance of the SteelHead package.

VM Type specifies Other from the drop-down menu. You must select Other.

Dedicated Cores (Optimize) specifies Yes from the drop-down menu.

Serial Console specifies Enable from the drop-down menu.

Sriov Driver(s) enables you to select all available options if you plan to use SR-IOV on the primary and auxiliary interfaces.

Raw Disk File Bus specifies Virtio from the drop-down menu.

Thick Disk Provisioning specifies Yes if you are deploying to a production environment.

Accept the default values for items in the bootstrap section.

Select Raw Images (.qcow2/.img) and upload both of the qcow2 files you created in Preparing images without using scripts in this order: mgmt.qcow2 file and then data store qcow2 file.

Optionally, you can create preconfigured deployment profiles using the Advanced Configuration settings.

After the qcow2 files are uploaded, submit them. The uploaded files are packaged into a tar.gz file, and then the tar.gz file is added to the list of packages at the bottom of the Image Packaging tab.

Register the new package. Registered images can be used on any NFVIS system.

Before you deploy the appliance, ensure that the virtual environment has:

Some elements are created for you by the system using default values, but you will need to manually create the router and service-net virtual switch.

After your virtual environment is in place, you can create the appliance, assign interfaces, and then deploy the environment including the appliance.

Remove the connection between the GEO-3 port from the lan-net virtual switch. To do so, choose VM Life Cycle > Networking. In the Networks & Bridges section, find the row for the lan-net virtual switch, and click the edit icon (blue pencil). In the lan-net virtual switch details page, find the Interfaces field and remote GEO-3, and then click Submit.

Choose VM Life Cycle > Deploy. Drag and drop an Other icon from the palette at the top of the VM Deployment page to an open space in the canvas below. Ensure the Other icon on the canvas is selected, and then under VM Details specify these items:

VM Name specifies the name for the SteelHead.

Image specifies the .tar.gz image.

Profile selects the profile.

Deployment Disk specifies Internal. For the Cisco 5100 Series ENCS, this item must be set to Internal.

One at a time, drag and drop NETWORK icons onto the canvas, connecting one end of each to the appliance and the other end to the lan-net virtual switch.

Deploy the setup. Deployment is complete when the status of the virtual machine changes from Deploying to Active.

Start the appliance. Startup is complete when the console displays a login prompt.

The deployment package for Nutanix is the same TAR archive used for Linux.

After you have downloaded the package and extracted its contents, upload the mgmt.qcow2 image file to your Nutanix Prism system. After upload, confirm that the image is available and that its status is Active.

After upload and confirmation of its availability, create a virtual machine capable of hosting the model you want.

When adding a management disk, use Clone from Image Service from the Operation, select PCI as the bus type, and clone the image from the deployment package. Ensure that your newly added disk is first in boot priority. Depending on the appliance model, you may need to expand the disk. For details, go to Knowledge Base article S29147.

When adding a data store disk, select PCI as the type and allocate it on the storage container.

You need to add a total of four NICs, one each for these interfaces: primary, auxiliary, LAN, and WAN.

Power on the virtual machine, and then log in to the appliance and complete the initial configuration.

Auto bootup and keystroke entry on the Nutanix default console has known issues. Perform the following actions before trying to access the virtual machine’s console: