Guest post from Lori Mac Vittie (@lmacvittie) from F5 Networks
How you provision all the network things matters
Polymorphism is a concept central to object-oriented programming. The notion of polymorphism is used to extend the capabilities of a basic object, like a mammal, to specific implementations, like cats or dogs or honey badgers, even though they don’t care about such technical distinctions. A good example of this is cats and dogs, which are both of the type “mammal” but that “speak” in a different voice.
This becomes important as we consider the way in which Cisco Application Policy Infrastructure Controller (APIC) enables the extension of automation across the network, particularly to the application layers (L4-7), because it deviates from traditional protocol-based methods as a way to facilitate and automation service insertion in a common way without limiting the robust capabilities of best-of-breed solutions. In other words, it doesn’t require all mammals to speak in the same voice.
Traditional protocol-based methods rely on a common data model. A TCP packet, for example, contains a specific set of headers that describe a variety of options and characteristics of the flow. The specific format is prescribed by RFCs and no deviation is allowed. Network integration has generally followed this model and you can see the results in a variety of ongoing efforts to provide orchestration and automation across the network. All devices are treated like mammals. There are no dogs, there are no cats, and there are certainly no honey badgers. The result is a commoditized set of network capabilities which do not allow the differentiation in services or enable the per-application attention required to address application-specific challenges with security, performance and scalability.
Which brings us back to Cisco APIC and its Application Centric Infrastructure (ACI) approach, which lets honey badgers be honey badgers and cats be cats while still both being mammals.
The Cisco ACI approach is very object-oriented. Its integration model requires the existence of a set of functions, but in no way prescribes how those functions act. This means that a variety of solutions in the same market can all integrate with Cisco APIC, but any capabilities that go above and beyond the lowest common denominator are not lost. Because of the dynamic nature of the integration – via device packages that can be loaded at any time – that also means that integrations can continue to be developed that enable even greater ranges of flexibility and choices for customers. In other words, you aren’t stuck with just cats, dogs or honey badgers. You can also bring in guinea pigs, rabbits and horses, as long as they’re mammals and each implements the basic set of functions required of a device package deployable on Cisco APIC.
It is that extensibility that has enabled F5 to continue to expand the choices available for integrating the automation of L4-7 application service insertion with Cisco APIC. Initially our focus was on direct integration with BIG-IP, providing the means by which prescriptive provisioning and configuration of services was easily accomplished. But the reality is that the applications central to driving the application economy are not all one size fits all. An approach that enables more specific, per-application service provisioning is necessary to achieve the operationalization of app deployments needed to relieve the increasing pressure faced by 9 out of 10 executives to release apps more quickly (CA and Vanson Bourne, Global Application Economy Study 2014).
Because of the approach Cisco has taken to enabling that provisioning via Cisco APIC, F5 Networks is able to provide another integration path through its orchestration and management solution, BIG-IQ. This new integration option facilitates the use of per-app service templates, iApps, to ensure not only rapid deployment but custom and consistent configuration. Consistency is an important capability necessary for maintaining stability in an infrastructure ultimately responsible for delivering the hundreds of applications supported by the average enterprise that must be balanced against the need for faster, more frequent deployments. Customization is required by the very concept of application-centricity, as no two applications are alike in terms of the services and characteristics of those services required to meet business and customer expectations.
This application-focused approach to provisioning allows network and application operators alike to codify per-application service requirements along with common policy such as base security using an app template approach. These templates then become the core of a custom device package that can be loaded and executed via Cisco APIC, resulting in a rapid, consistent deployment of the app services critical to ensuring the performance, security and scalability of the applications driving the application economy.
We are as excited today with the introduction of our BIG-IQ integration with Cisco APIC as we were with our BIG-IP integration. We’re particularly pleased with Cisco’s model of integration precisely because it enables us to continue to protect our customers’ investment in the technologies and capabilities that go above and beyond the basics when it comes to delivering application services.
You can experience the integration between Cisco APIC and both F5 BIG-IQ and BIG-IP in a live demonstration covering design and deployment options, troubleshooting and monitoring as well as customer use case scenarios in our upcoming SDx Central Demo Friday, March 13th at 10am PST
We also have workshops for you to attend in case you happen to be in Toronto March 24, Montreal March 25 or Ottawa March 26. Register now for the Cisco and F5 Synthesis Workshop: Accelerating Application Deployments.
Tags: Cisco APIC, F5, OO, SDN
Big data and analytics are changing the competitive playing field. Winners will find ways to turn unprecedented growth and complexity of big data, cloud, IoT devices and more into competitive advantage.
The Value of Big Data Integration
Big data itself can provide significant insight, impact and advantage. These outcomes can be multiplied when big data is combined with existing enterprise data or even other big data. These insights can lead to enriched customer engagement, faster new product innovation, higher returns on capital and more. The possibilities are endless. And your competition has the same objectives.
As a result, integrating big data intelligently is a challenge on nearly everyone’s short list. Three patterns of integration have formed around big data integration.
Cisco Data Virtualization is proving a key enabler, delivering many of the capabilities required. With Cisco UCS and key partners including Appfluent, Cloudera, Hortonworks, Informatica and MapR providing additional value add.
Integration Pattern 1 – Hadoop to Hadoop Integration
The Cisco Hadoop-to-Hadoop Data Integration solution provides a more agile and faster way to access and query multiple Hadoop clusters. This solution enables even “bigger” big data analytics for greater insights and business impacts.
Integration Pattern 2 – Hadoop to Enterprise Integration
The Cisco Hadoop-to-Enterprise Data Integration solution provides a more agile and faster way to bring together traditional enterprise data with the new world of big data. Accessing and querying both Hadoop and enterprise data enriches the analytic opportunities and results in better business decisions.
Integration Pattern 3 – Data Warehouse Optimization Integration
With data warehouse data doubling every two years, data warehouse costs are skyrocketing. To reduce these ever-growing warehouse management costs Cisco and Informatica recently announced a joint Data Warehouse Optimization solution.
This offering combines the hardware, software and services required to successfully offload processing and storage from data warehouses to Hadoop. Not only does it significantly reduce costs, the solution also help organizations integrate and analyze more data and types of data, and thus drive greater business impacts.
At Strata + Hadoop World last week, I discussed this new solution with Jim McHugh, VP of UCS & Data Center Solutions Marketing at Cisco.
To learn more about these solutions and how Cisco’s rich portfolio of big data and analytics solutions can help you unlock the value of your big data, check out these two on-demand seminars:
Unlock Your Competitive Edge with Cisco Big Data and Analytics Solutions
Analytics Solutions for Driving Better Business Outcomes
Join the Conversation
Follow us @CiscoDataVirt.
Learn More from My Colleagues
Check out the blogs of Mala Anand, Mike Flannagan and Nicola Villa to learn more.
Tags: Big Data, data and analytics, data integration, data virtualization, data warehouse optimization, Hadoop, UCS
Cisco Intelligent Traffic Director (ITD) is a zero latency multi-terabit layer 4 load-balancer available on 5k/6k/7k/9k. It has support for traffic steering and clustering solution on the Nexus series of switches.
ITD allows customers to deploy servers and appliances from any vendor with no network or topology changes. With a few simple configuration steps on a Cisco Nexus switches, customers can create an appliance or server cluster and deploy multiple devices to scale service capacity with ease. The servers or appliances do not have to be directly connected to the Cisco Nexus switch. ITD takes load balancing decision in a single clock cycle of the switching hardware, Hence avoiding latency altogether.
ITD supports IP-stickiness, resiliency, NAT (EFT), VIP, health monitoring, sophisticated failure handling policies, N+M redundancy, IPv4, IPv6, VRF, weighted load-balancing, bi-directional flow-coherency, and IPSLA probes including DNS. There is no service module or external appliance needed.
ITD provides order of magnitude CAPEX and OPEX savings for the customers. ITD is available on Nexus 7000/7700 series in NX-OS 6.2(8) or later. It is available for demo/EFT on Nexus 5k/6k/9k. ITD is much superior than legacy solutions like PBR, WCCP, ECMP, port-channel, layer-4 load-balancer appliances.
Note:- ITD is not a replacement for L7 load balancers
Tags: Cisco Nexus 9000, ITD, latency, load balancer, nexus, Nexus 7000 Series Switches, Nexus 9000 Series Switches, performance
Many years ago I found myself talking to venture capitalists about the differences between SaaS, outsourcing, ASPs, MSPs, online applications; etc. Also I noticed that my Stanford students had little understanding of the economics of software, so I developed the idea of seven business models to cover everything in the software business, and remove the buzzwords and replace them with economic models.
In my previous blog post we discussed the first four models, this post will cover Models Five through Seven.
We ended the last blog talking about Model Four being able to provide management of the security, availability, performance and change of the software at nearly 10x less cost.
The question we left with was “how”?
How is it possible to decrease the cost of management without just paying people a fraction of what they made previously?
Read More »
Tags: business models, CIO, Cloud Computing, isv, on-premise, open source, outsourcing, SaaS
It’s been a busy couple of weeks for us in big data land. One thing that struck me is how much we learned about the big data and analytics space after hours at Strata Hadoop, and sometimes that can be just as exciting as what we learn during the sessions. I have a couple of videos I like to share with you to prove my point.
In this video, I learned some things from Jim Scott, MapR. First, MapR is providing –and get ready because it’s a mouthful—“ an online, Interactive, platform neutral, vendor agnostic training” — and some cool use cases of a Hadoop cluster in a briefcase.
It’s also fun to get together with our various partners and talk about the great solutions we are doing together. In this video, I have MapR and Platfora do all the work. Hey it was after hours!
Webinar At A Glance
Want to learn more about what we are doing with our partners? Check out the Webinar At A Glance from our recent Big Data webcast, “Analytics Solutions for Driving Better Business Outcome,” which is available on demand now. The feedback has been amazing and we plan to do more of these with our partners in the future.
Have some interesting stories or solutions you would like to share? Find me on Twitter, @JimMcHugh, we can work together to get the news out.
Tags: analytics, Big Data, BigData, CiscoUCS, MapR, Platfora, Splunk, Strata Hadoop World, unified computing system