Using Cloud Accelerator for VMware ESX/ESXi

VMware ESX/ESXi is hypervisor software that enables the creation and management of virtual machines. Virtual machines act as hosts for Cloud Accelerator virtual appliances. Installing a Cloud Accelerator image on a virtual machine is much like installing RiOS software on physical Riverbed hardware.

The Cloud Portal supports licensing and basic status reporting for ESX/ESXi-hosted
Cloud Accelerator virtual appliances.

Cloud Accelerator for ESX/ESXi Limitations

Cloud Accelerator for ESX/ESXi has these limitations:

• Provisioning—You must provision ESX resources manually. You can use any provisioning tool provided by VMware-based cloud providers such as vSphere.

• Virtual machine operations—You must use the cloud provider toolset (such as the vSphere tools
or Cloud Console) to start, stop, pause, and resume the virtual machine.

• Discovery—The portal does not provide discovery services for Cloud Accelerator, because
a Cloud Accelerator deployed on an ESX cloud requires static IP addresses to work correctly.

Cloud Accelerator for ESX/ESXi requirements

This table describes the Cloud Accelerator for ESX/ESXi requirements.

Component	Requirement
Supported vSphere releases	4.0 and later.
Provisioning	OVA package deployment.
License management	Connectivity between Cloud Accelerator and the Cloud Accelerator (TCP port 80 connection to cloudportal.riverbed.com).
Virtual hardware	64 bit only. vCPU: One 1.2 vCPU to four 2-GHz vCPUs Reserved RAM: 2 GB to 6 GB Disk capacity: 470 GB
IP addressing	Static, when you use the Discovery Agent. Cloud Accelerator does not support accessing client-side SteelHeads with overlapping IP addresses.
Firewall setting	Disabled (if the rules exclude the TCP option) or modified target rules. Enable port 80 to access cloudportal.riverbed.com. Cloud Accelerator supports only traffic in clear text or SSL encryption. It does not support IPsec encryption.
Discovery Agent	Supported Windows versions: Windows 2003, Windows 2008 R1, and Windows 7. Supported Linux versions: CentOS 5.0, 5.2, 5.3, 5.4, 6.3, 7.3, and 7.5 (32 bit and 64 bit), Linux Ubuntu 8.04 and 10.04, and RedHat 4 and 5.
Server VM limitations	Only supports server VMs with a single virtual NIC.

Basic steps to deploy a Cloud Accelerator for ESX/ESXi

This table lists the deployment tasks.

Task	Reference
1. Install the Cloud Accelerator using VMware vSphere.	Installing the Cloud Accelerator virtual appliance for ESX/ESXi
2. Configure the resources that the ESX server will be granting the Cloud Accelerator.	Configuring ESX resources
3. Complete the initial configuration of the Cloud Accelerator.	Completing the initial configuration
4. Obtain the one-time token from the Cloud Portal.	About Licensing Cloud Accelerator virtual appliances
5. Apply the one-time token to the Cloud Accelerator.	About Licensing Cloud Accelerator virtual appliances
6. Use the Cloud Portal to monitor the Cloud Accelerator.	About the Cloud Portal
7. If you decide not to use the Discovery Agent, go to Step 9.
8. Install and manually configure the Discovery Agent on the server. You cannot use the Cloud Portal to configure the Discovery Agent for Cloud Accelerator. You must use the manual mode.	About the Discovery Agent
9. If you are not using the Discovery Agent, configure the client-side SteelHead in-path rules to point to the Cloud Accelerator for server.	SteelHead User Guide
10. If your network is NATed, go to Step 11.
11. If you are not using the Discovery Agent, configure the Cloud Accelerator for public or private IP address mapping.

Installing the Cloud Accelerator virtual appliance for ESX/ESXi

Riverbed provides the Cloud Accelerator for ESX/ESXi as an image that contains the VMX and VMDK files necessary to create the virtual machine.

The Cloud Accelerator image is an installable Open Virtual Appliance (OVA) package. OVA is a platform-independent, efficient, extensible, and open packaging distribution format. The OVA package provides a complete specification of the Cloud Accelerator, including its virtual disks, CPU, memory, networking, and storage. To install a Cloud Accelerator model other than the base model, first install the base model and then upgrade it to a higher model.

The OVA is a compressed package that quickly creates a virtual machine with predefined settings.

To obtain the OVA package, log in to your customer account at https://support.riverbed.com.

Each package contains predefined hardware requirements and configuration for the base model
Cloud Accelerator. Do not open or modify any of the files in the package. The package files take up several gigabytes of disk space (the package itself is less than 1 GB).

A Cloud Accelerator for ESX/ESXi requires a 440-GB virtual disk. This size is bigger than the 256 GB maximum virtual disk size deployable in an ESX cluster when you use the Virtual Machine File System (VMFS) default block size of 1 MB. Therefore, before installing a Cloud Accelerator, ensure that the target data store (VMFS) has a block size greater than 1 MB. This limitation is not valid if you use Network File System (NFS).

To install a Cloud Accelerator for ESX/ESXi

1. Obtain the VM package from https://support.riverbed.com and download it locally.

2. Extract the contents of the OVA package using the tar -xvf command or a freeware application such as 7-zip.

3. Open VMware vSphere, enter the hostname or host IP address, enter your username and password, and click Login.

4. Choose File > Deploy OVF template.

5. Select Deploy from file, click Browse, select the OVA file, and click Open.

6. Click Next to display the OVF Template Details page.

7. Verify that the OVA file is the one you want to deploy, and click Next to display the Name and Location page.

8. Enter a name for the virtual machine.

9. Click Next to display the Datastore page.

10. Select a data store in which to store the virtual machine and its virtual disk files. Here are some tips about the data store:

– The standard installation puts both VMDKs on a single data store.

– The data store holds the virtual machine files and is not used for the Riverbed Optimization System (RiOS) data store.

– Make sure the data store you select has enough capacity for the OVA package to install.

– We recommend that you put the larger VMDK containing the RiOS data store on the fastest available data store. The data store should have enough room to expand to the required size of the Cloud Accelerator model.

– The smaller VMDK containing the management system can be installed on any data store type.

– Do not share host physical disks (such as SCSI or SATA disks) between VMs. Select an unshared disk for the data store disk.

– Do not delete data store disk 1 (DS1).

11. Click Next to display the Disk Format page.

12. On the Disk Format page, select Thick provisioned format.

Thick provisioning preallocates all storage.

13. Click Next to display the Network Mapping page.

14. Select the destination network name and choose a network from the drop-down list to map the source network to a destination network.

Ensure that the LAN and WAN interfaces (NIC3 and NIC4 in ESXi) are not connected to the same virtual switch. Otherwise, the system displays the error message:

Failed to apply configuration change(s) Internal error: module commit apply function of the module rbt.

Make sure that you map each source network to a unique destination network. If a source network is mapped to the same destination as another source, an error message appears. Mapping source networks to the same destination network can create a loop in the system and might make your ESX host unreachable.

15. Click Next to display the Ready to Complete page.

16. Verify the deployment settings and click Finish.

A dialog box shows the amount of time it will take for the deployment to complete.

When the deployment finishes, a dialog box tells you that the deployment was successful.

17. Click Close.

The new virtual machine appears under the hostname or host IP address to the virtual machine inventory.

Configuring ESX resources

Before you power on the Cloud Accelerator, you must configure the resources that the ESX server will be granting the Cloud Accelerator.

To configure ESX resources

1. Right-click the virtual machine you created and click Edit Settings to display the Virtual Machine Properties page.

2. Select the Hardware tab and click Memory in the Hardware column.

3. Increase the memory size to at least the minimum required by the model you want to use.

4. Click OK to save your changes.

5. Click Edit Settings and select the Hardware tab. Click CPUs in the Hardware column.

6. Set the number of virtual processors to at least the minimum required for the model you want to use.

7. Click OK to save your changes.

8. Click Edit Settings, select the Resources tab, and select CPU.

9. Under Resource Allocation, change the Reservation to at least the minimum required for the model you want to use.

10. Click OK to save your changes.

11. Select the virtual machine and choose Power > Power On.

12. Select the Console tab.

The Cloud Accelerator starts and the login prompt appears.

13. Log in to the Cloud Accelerator using the default login admin and the default password password.

Completing the initial configuration

This section describes how to complete the initial configuration of the Cloud Accelerator for ESX/ESXi.

Press Enter to enter the default value. If you mistakenly answer no, you can start the configuration wizard by entering configuration jump-start at the system prompt.

Press ? for help. Press Ctrl+B to go back to the previous step.

To configure the Cloud Accelerator for ESX/ESXi

1. After you log in to the Cloud Accelerator as administrator, the system prompts you to start the configuration wizard.

2. Complete the configuration wizard steps on the client-side Cloud Accelerator as described in this table.

Wizard prompt	Description
Step 1: Hostname?	Enter the hostname for the SteelHead.
Step 2: Use DHCP on the primary interface?	You are given the option to enable the DHCP to automatically assign an IP address to the primary interface for the Cloud Accelerator. We recommend that you do not set DHCP. The default value is no.
Step 3: Primary IP address?	Enter the IP address for the Cloud Accelerator.
Step 4: Netmask?	Enter the netmask address.
Step 5: Default gateway?	Enter the default gateway for the SteelHead.
Step 6: Primary DNS server?	Enter the primary DNS server IP address.
Step 7: Domain name?	Enter the domain name for the network in which the Cloud Accelerator is to reside. If you set a domain name, you can enter hostnames in the system without the domain name.
Step 8: Admin password?	We strongly recommend that you change the default administrator password. The password must be a minimum of six characters.
Step 9: SMTP server?	Enter the SMTP server name. External DNS and external access for SMTP traffic is required for email notification of events and failures to function. Make sure that you provide a valid SMTP server to ensure email notifications for events and failures.
Step 10: Notification email address?	Enter a valid email address to which notification of events and failures are to be sent.
Step 11: Set the primary interface speed?	Enter the speed on the primary interface (the Cloud Accelerator). Make sure that this value matches the settings on your router or switch. The default value is auto.
Step 12: Set the primary interface duplex?	Enter the duplex mode on the primary interface. Make sure that this value matches the settings on your router or switch. The default value is auto.

3. The system confirms your settings. To change an answer, enter the step number to return to; otherwise, press Enter to save changes and exit.

4. To log out of the system, enter the exit command.

Logging in to the Cloud Accelerator Management Console

This section describes how to log in to the Cloud Accelerator Management Console. The Management Console makes managing the Cloud Accelerator simpler through a web browser interface.

You can connect to the Cloud Accelerator through any supported web browser. To connect to the
Cloud Accelerator, you must know the hostname and domain, or the IP address, of the virtual appliance.

For first-time login, the default username is admin and the default password is password. If you specified administrator credentials at the time you created the virtual appliance instance, use those credentials.

Cookies and JavaScript must be enabled in your browser.

To log in to the management console

1. Enter the URL for the Cloud Accelerator in the location box of your browser:

<protocol> is HTTP or HTTPS. The secure HTTPS uses the SSL protocol to ensure a secure environment. If you use HTTPS to connect, you are prompted to inspect and verify the SSL key.

<host> is the IP address or hostname you assigned to the Cloud Accelerator during the initial configuration. If your DNS server maps the IP address to a name, you can specify the DNS name.

<domain> is the full domain name for the Cloud Accelerator.

Alternatively, you can specify the IP address instead of the host and domain.

The Management Console Login page appears.

2. In the Username text box, enter the user login: admin or monitor. The default login is admin.

Users with administrator privileges can configure and administer the SteelHead. Users with monitor privileges can view connected SteelHeads and reports. A monitor user cannot make configuration changes, modify private keys, view logs, or manage cryptographic modules in the system.

3. In the Password text box, enter the default password or the one you assigned when you created the virtual appliance instance.

4. Click Log In to display the Home page.