It’s kinda funny how things go full circle. today, at AWS re: Invent, we launched NetScaler on AWS. Literally. As in, from the AWS Marketplace, you can launch a NetScaler AMI or a CloudBridge AMI into your AWS cloud environment.
NetScaler got its start in the cloud, before we were even calling it the cloud. That is, NetScaler’s first deployments were in the data centers that hosted the first wave of super-massive “webscale” applications. These were the first cloud data centers. The design principles forged in these data centers over time have evolved into the core architecture used in today’s largest public and private clouds.
the trial-by-fire of being “in-line” in these data centers influences everything we do. It drives a lot of our functional enhancements (including things you wouldn’t normally associate with an ADC like Datastream database load balancing and AppFlow for application visibility). The elasticity requirements of these cloud data centers drove the thinking behind NetScaler TriScale. And it’s the reason why, more than two years ago, we started working with AWS to make NetScaler available as an AMI on AWS.
Fundamental to this effort is that “NetScaler needs to be NetScaler.” The AMI is a packaging of the same binary used on NetScaler MPX and SDX appliances, and NetScaler VPX virtual appliances. NetScaler in AWS supports all the same L4-7 functionality (e.g., load balancing, content switching, global server load balancing, SSL VPN, application firewall, etc.) that runs on-premise. This is another example of things coming full circle. We take for granted we can get all of the above in a single device on-premise. Until now, in the cloud, we’ve had to knit together separate VMs for L4 load balancing, L7 manipulation, caching, security, etc. NetScaler on AWS allows you consolidate all this into a single AMI, all governed by the same policy management framework. Using the recently introduced elastic network interface (ENI) and multi-IP functionality of AWS VPC, a single NetScaler instance can host multiple VIPs, multiple vServers, and attach to various subnets within a VPC. Without functional and configuration parity, all the use cases (dev/test, DR/BC, production spillover, or even just simple migrating an app or workload from on-premise to the cloud) that involve spanning own data centers and the cloud break.
Also, fundamental to getting NetScaler on AWS was supporting the elasticity/flexibility we expect of the cloud. From a technical perspective CloudBridge on AWS really helps here, as it makes it real easy to set up a WAN-optimized, secure tunnel between your data center and AWS. This in turn helps make AWS look like a natural extension of your enterprise network. Global server load balancing is going to play a huge role here as well, for both scaling out capacity, and leveling load/distributing traffic between and across AWS zones, AWS regions and on-premise data centers.
To further provide elasticity, in addition to enabling “bring your own license” (including all the pay/grow and bursting you’ve come to expect), NetScaler is available “by-the-hour” on the AWS Marketplace. Whether you need a single NetScaler for a short period of time, a LOT more NetScalers for a period of time, or you just need a NetScaler right now, AWS Marketplace allows you to launch it/them immediately, and then spin them down just as immediately. That’s cloud elasticity; you pay only for what you use when you use it.
If you’re at AWS re: Invent, feel free to swing by the Citrix booth. We can walk you through all the details (though again, NetScaler is NetScaler). Some of our lead engineers and architects are here as well. They, like you, will be attending a lot of the sessions, but they’ll also be at the booth from time-to-time as well.