The NS VPX is available as an AMI in AWS marketplace, and it can be launched as an EC2 instance within an AWS VPC. The VPX AMI instance requires a minimum of 2 virtual CPUs and 2 GB of memory. An EC2 instance launched within an AWS VPC can also provide the multiple interfaces, multiple IP addresses per interface, and public and private IP addresses needed for VPX conﬁguration. Each VPX instance requires at least three IP subnets:
- A NetScaler Management subnet
- A client-facing subnet (VIP)
- A back-end facing subnet (SNIP,MIP, etc.)
Citrix recommends three network interfaces for a standard VPX on AWS installation.
AWS currently makes multi-IP functionality available only to instances running within an AWS VPC. A VPX instance in a VPC can be used to load balance servers running in EC2 instances. An Amazon VPC allows you to create and control a virtual networking environment, including your own IP address range, subnets, route tables, and network gateways.
Note: By default, you can create up to 5 VPC instances per AWS region for each AWS account. You can request higher VPC limits by submitting Amazon's request form (http://aws.amazon.com/contact-us/vpc-request/).
Figure 1. A Sample NetScaler VPX Instance Deployment on AWS Architecture
Figure 1 shows a simple topology of an AWS VPC with a NetScaler
VPX deployment. The AWS VPC has:
- A single Internet gateway to route traffic in and out of the VPC.
- Network connectivity between the Internet gateway and the Internet.
- Three subnets, one each for management, client, and server.
- Network connectivity between the Internet gateway and the two subnets (management and client).
- A standalone NetScaler VPX instance deployed within the VPC. The VPX instance has three ENIs, one attached to each subnet.