Hello World! This is my first post on the Riverbed blog. So let me start it by asking you a question "How many times has someone asked you about your cloud strategy over the past three months?" Is it every day, every other day, a few times, or not at all? If this were a slashdot poll I bet the "not at all" option would be running last. It seems clouds are every where (they certainly are in Cambridge, UK this morning). So what is your cloud strategy? Hah. Now you have to change your answer!
Regardless of how you currently stand on "The Cloud" or where you strategy will take over the next few months, Riverbed can help. We can help in Physical Data Centres, we can help in Virtual Environments, and we can help in the cloud. So even if your cloud strategy is "No thank you", we can help. But as you're still reading a post titled "Cloud Bursting..." lets suppose that you're atleast a little curious about this cloud thing.
So now I could start talking about the Riverbed Cloud Steelhead and how it can help accelerate your data between the cloud and you physical DC or branch office. Alternatively I could start waxing lyrical about the SaaS accelerator we've recently launched with our good friends Akamai. But I'm not going to talk about either of those. The title of this post says Stingray, and Stingray is what you're going to get. Stingray Traffic Manager and Cloud-Bursting specifically.
So What is Cloud Bursting?
Cloud-Bursting is a deployment model which allows you to make use of cloud infrastructure on your own terms. It's often a hybrid cloud model where you continue to use your current physical or virtual Data Centre, and simply utilize resource within a third party cloud on demand. We're effectively taking dynamic infrastructure to it's logical conclusion; Data Centre on Demand. Hang on, that sounds complicated... How does it work?
An important component in bringing up Data Centres on Demand is the ability to bring up individual servers on demand. Fortunately Stingray TM allows you to do just that with the Auto-Scaling module. This can be used in Virtual or cloud environments with or without cloud bursting. Sometimes you just want to scale up within your Data Centre, Stingray Auto-Scaling allows you to do that. It's an important feature in cloud environments where you are billed per minute for running services. Stingray will ensure that you only have the servers running required to meet demand.
Global Load Balancing
Another important component in our Data Centre on Demand strategy is a Global Load Balancer (GLB). We need a GLB component to detect when new Data Centres are spun up and to start directing traffic over to them, once they are ready to receive that traffic. Obviously the GLB component needs to include sophisticated monitoring to achieve this. Guess what? Stingray Traffic Manager has this too.
Event Mapping (Orchestration)
Thirdly we need some way to detect the need for a new Data Centre and trigger the deployment process. Stingrays sophisticated Alerting and Event system allows us to trigger the Cloud-Bursting based on any number of events. We could trigger it from Stingrays scripting language TrafficScript, we could use a local Auto-Scaling limit, or Service Level Monitoring thresholds. We could even base it on how much money your website is turning over per minute. Extensibility is Stingrays middle name.
Show me the demo
So this all sounds very good, but I'll believe it when I see it. Right? Well then, let me show it to you..