F5 Application Services Reference Architecture (Audio)

Editor's Notes

• #2: Welcome to the F5 Application Services Reference Architecture
• #3: Discuss this quote:- Changing environment? Flexible? Needs of the customers? Why is it changing, what are the needs of the customer?
• #4: Users want three simple things from applications, always on always fast and always available. How have we got here? It’s driven by consumer applications – like Pandora Music (available in my car, on my phone, pc and TV), like facebook, like online banking systems. We have grown to expect that our personal apps work like this – and we expect all out apps to work like this now. And guess what – if they don’t we can go elsewhere – even if we are a business unit of a large corp (reference jive, attask, sdfc) etc.
• #5: So it’s a difficult world – there are more mobile devices than atoms in the universe than ever, applications and their users are changing faster and faster Users and businesses can choose dozens of solutions form in-house to cloud to Saas apps If the business is going to stay relevant to its customers or the IT department to the business then they need to provide the same things
• #6: Why is the rate so high? Sometimes it’s poor project management, sometimes it’s a poorly thought through system. Mostly its because what was delivered wasn’t what was needed – it’s down to poor alignment between IT and business. One of the ways organizations are attempting to keep projects on track is by brining Agile development methods to application roll out and update Instead of massive, infrequent updates organizations are delivering smaller, more frequent changes To do this tight integration between application development teams and operations teams is a necessity DevOps is the result of this
• #7: Another complication is that Linear delivery - where applications live in corporate datacenters and are delivered to PC based users is gone
• #8: Now, we have a complex matrix of application sources being delivered to a more dispersed, more diverse population. That adds another set of challenges. [If there is time and you’re feeling loquacious, then expand on this]
• #9: Meet user expectations Accelerate your deployment cycles Without sacrificing stability Or increasing complexity? Or cost. It’s tough
• #10: One place you can get help is the application delivery layer
• #11: At the core of what we do is still the functionality we bring: load balancing, availability, security, and acceleration. But on their own, these functions are not enough. We need to give you more. We must surround this core with ways to help you get the right services to the right application at the right point in the network. We can do this with three things:- Programmability, fluency, and scalability. Because when you add programmability, fluency, and scalability to this core functionality, then you start to create the kind of scalable, integrated application delivery fabric that can help you solve these challenges.
• #12: At the heart of what we do is the huge array of functionality we can bring to applications and networks. And it’s such a rich and deep vein that we can only scratch the surface here. Each of these buckets of functionality contains an array of repeatable functions like relocation, load balancing, data scrubbing, or compression. Using the services we can provide helps you to adjust the balance of bespoke coding or configuration you use to model where reusable, repeatable services are delivered from the network.
• #13: But how do you know what services to deliver – and how can a network supply services to an APPLICATION? Well, the first thing we need to do to add application-layer value is to understand the application. We’ve built more than 10 application specific proxies, and there are more on the way. The most obvious might be HTTP but we have SIP, FTP, FIX etc. When you understand what’s going on at the application layer, you can add application-specific optimizations or policies. But how do you know what to deliver? There is so much you could do – but what should you do? We have done the work to determine this and then made it incredibly easy for you to access our expertise in the form of iApp templates, which provide not just a simple way to deploy applications but also a way to manage and view their component.s This brings applications and networks together to form application networks.
• #14: Programmability. Why? Well, because it lets you innovate and integrate your infrastructure. Programming the data plane with iRules lets you create your own functionality, or adapt existing features to your needs. Programming the control plane means you can reconfigure the system in response to changes in application traffic or external events. Finally, programming the management plane allows you to connect your network to your deployment or orchestration systems.
• #15: The final part of our story is scalability. Scale up with VIPRION and license upgrades. Scale out with ScaleN up to 32 nodes. Scale into the cloud with the same feature sets and programmability as other platforms – and build secured pathways. Change isolated pockets of capacity and capability into a complete delivery fabric.
• #17: To accelerate the application you need to first understand the context – what is the client? What is the latency? What kind of request are they making? Which service will they use? Then you can go about optimizing their network connectivity – based on the information you know. Mobile works best with different TCP settings from desktops. Finally, you can apply application-, network-, and device-specific optimizations to the traffic -- don’t heavily compress objects to local devices, for instance.
• #18: Now you can use F5 Application Services to accelerate your development and deployment cycles. Build apps with common network services. Test them on safe environments, but with the same technology. Test them on a fraction of production. Deploy them automatically.
• #19: More users? Increase resources: scale up, out, or across. Servers overloaded? Offload, limit, or redirect. Orchestration inputs. Infrastructure failures. Feed back data about performance into monitoring systems.
• #26: F5 offers three enterprise offerings to consolidate advanced traffic management, optimization, and security services on one BIG-IP platform. NOTE: We will not remove any SKUs; a la carte SKUs and single module offerings are still available to all customers. Good, Better, Best offerings are a simplification overlay that removes complexity without removing the current levels of agility. Flexibility: Leverage F5 innovative and advanced technologies across your infrastructure, meeting the needs of the most demanding, mission-critical applications Simplicity: Consolidate application development, deployment and management requirements into fewer common configurations Value: Save up to 65% lower prices vs. buying as components Alternate: Flexibility: Leverage F5 technology across your infrastructure Simplicity: Consolidate into fewer common configurations Best Pricing: Up to 65% lower prices vs. buying as components