We use cookies and other similar technologies (Cookies) to enhance your experience and to provide you with relevant content and ads. By using our website, you are agreeing to the use of Cookies. You can change your settings at any time. Cookie Policy.

Cloud Computing Insights

Tim Vibbert

Podcast session with IBM

Vote 0 Votes

On March 30 I conducted the second session of the Cloud QCamp podcast series with Snehal Antani Lead Strategist for Websphere Cloud and Product Management.  Here is the audio and transcript from that session. 

Download file

Announcer: Welcome to another ebizQ podcast.

TV: Híŋhaŋni wašté. (Good day) and welcome to the Cloud QCamp podcast series. I'm your host Tim Vibbert. This podcast series is being conducted by Red Eagle Services in conjunction with the ebizQ Cloud QCamp Virtual Conference being held 7 April. I would like to welcome my guest today, Snehal Antani, Lead Strategist and Architect at IBM. Welcome Snehal.

SA: Thanks.

TV: IBM has a vast cloud computing strategy ranging from platforms and infrastructure to development and analytics. The IBM strategy covered both public and private clouds. Snehal, I have a couple of questions I'd like ask. First question is what is cloud and how did we get here?

SA: Sure. So if you look at the characteristics of a cloud where we really think about this as the characteristics of an evolved middleware infrastructure. We want our middleware infrastructures to be responsive where resources can be allocated on the fly to meet demand. We want our infrastructures to be resilient where can reduce the probability of a failure, isolate the impact of those failures and recover very quickly from those failures.

We want to ensure that we're virtualized and maximizing system resources and lowering our overall ownership of the applications and the underlying technology, the level of management where they're self-service, significant levels of automation, and very much policy driven management approaches versus the traditional point and click and the automated, decentralized levels of manageability. And then finally, a level of agility where we're able to better align our IT capabilities with the business needs. And so these are the characteristics that we're trying to pursue. There's a number of ways to get -- there's a whole thought or methodology for how to get there.

And if you think about cloud as essentially falling into this notion of cloudonomics where we are taking our understanding and expertise within virtualization, within standardization, and automation and brining those together to reduce cost and increase flexibility. And cloud really brings these three sets of topics together in a more comprehensive solution where we're able to just build better middleware infrastructure, right. And so one of the things is how do we get here? Well, we think about the previous push within data centers where what we wanted to do is help out customers take their IT services themselves as business services and adhere to throw principles around decoupling, reuse, preparation of concerns, clearly define roles and responsibilities, use standards and interoperability to get different services to interact with one another.

We're able to then as a complement take those same principles that we learned from an application perspective and start to apply them to the infrastructure, the IT infrastructure itself where we decomposed the IT infrastructure from what was a conglomerate, holistic, very large monolithic middleware solution to infrastructure and Infrastructure-as-a-Service. Where we've got our hardware storage and networking, managed, virtualized, and served right as one extraction layer on which we've got our Platform-as-a-Service layer as a second extraction layer, where we've got our middleware, our databases, JAVA runtimes, Web 2.0 runtimes, whatever else, virtualized, [indecipherable] managed, automated, policy driven [indecipherable].

And then finally, the third layer -- the third level of extraction being the applications themselves and delivering software in and as a service models. So then here, we've got, for example, collaboration, or CRM, or industry specific solutions. And so we've evolved quite a bit, where we went from -- when we talk about evolution we talk about the evolution of the data center, to adhere to these types of principles and pursuing these types of middleware characteristics, or its responsiveness, resiliency of virtualization, manageability and agility.

TV: And you give a nice overview. As far as IBM goes, what is WebSphere's cloud strategy?

SA: Sure. So within the cloud strategy broadly for WebSphere, there are three pillars. So there is providing infrastructure for building clouds and then providing IBM middleware services in the cloud, in a public cloud, a private cloud and then providing connectivity among clouds. What we'll focus on here is the infrastructure for building clouds. So think about a private cloud example. Think of this as customers today what they would do if they had a collection of hardware, 25 boxes dedicated to your analytics workload, 50 boxes dedicated to your transactional batch, 30 boxes for OLTP, or whatever else.

It was very decentralized independent workloads, very siloed types of architectures and very siloed types of applications. And when we talk about building private clouds within the data center and thinking of this as a evolutionary thought, what we talk about is taking all this hardware, disparate hardware within a data center federating them into a single management domain where we're able to provide some level of server consolidation, improve utilization, centralize management, and then being able to carve out using workflow management technologies and autonomics, the different types of workloads running within this virtualized pool of hardware.

So then, we're able to dynamically determine how many resources should be allocated for our batch workloads, or our analytics workloads, or real time workloads, or OLTP workloads based on workload priority, application priority, resource consumption, demand for a particular application, so and so forth. And so in order to do that, a couple of things have to happen. So first, customers themselves have to evolve their data center where they move from siloed decentralized and dedicated world into a centralized and shared and highly efficient world. They pull that out pursuing things like standardization of their middleware, consolidation of their middleware and virtualization.

And then the next step thereafter is building out a private cloud or pursuing a private cloud that is responsive, resilient and agile. And to do that you got to focus on automation, intelligence, and providing more intelligence into the system, focusing on scaleability, and being able to hook into auto scaling types of features and so on and so forth. And so to talk about this in more detail terms, there are a number of things that customers have to consider in that evolutionary path. [Indecipherable] and private clouds, six pillars or so, the first is virtualization, automation, and cloud shaping where customers are going to have different sets of capabilities within that pillar and they got to recognize that if you want to build out a private cloud you got to be mature in your virtualization automation aspects of your middleware infrastructure.

And then there's a systems management pillar. There is cloud enabling your applications themselves to ensure that you are, for example, staple -- what applications are conducive to running within the cloud and how you manage [indecipherable] becomes interesting or for your batch applications ensuring that those batch applications run well within the cloud. And there you talk about writing, for example, single [indecipherable] code, that can be highly [indecipherable]. You got dynamic infrastructure, the ability to understand the demand for an application, and the autonomic responses to adjusting resources to meet that demand. Security services where you got to ensure that you're quite mature than your security domain of how you're doing your identity management, your single sign-on and your security propagation. And then connectivity service both within your data center among subsystems and across data centers or across clouds, we're talking about hybrid cloud connectivity.

TV: How might customers take advantage of clouds? You talk about evolving data centers, are there other advantages that customers can take and gain from say maybe a process sense?

SA: Right. What we should think about is customers today there's two very clear ways to leverage cloud and technology. The first is evolving their existing middleware infrastructure, to pursue those characteristics I talked about before, having responsive middleware infrastructure, resilient infrastructure, virtualized infrastructure for a lower cost, improved manageability so you'll lower your administrative costs and agility to better align IT capabilities with the business. And customers will leverage private clouds by pursuing those characteristics and this whole set of methodologies and technologies that they can pursue.

One thing that customers should understand is that it's not just a product thought. Products will help further enable or improve the capabilities of a customer's data center but there's also operational disciplines that they have to understand to take advantage of and change. There are application development disciplines that the customers need to understand and make changes to and adopt as well as expertise and services in how to build the right types of middleware architectures and designs. So that's the first part, evolving your middleware infrastructure.

And then the second part of that, there are elements of the application development lifecycle that are conductive to running within the public cloud for example. So think of this as transient workloads. So you may do your development on your laptop, for example, and your unit testing on your laptop but to do function testing, or to pursue system testing, or high scaleability testing where you got thousands of instances of your applications to ensure that you some level of linear scaleability. These are tests and elements of the app development lifecycle that require significant of resources or configuration and setup.

And its perhaps not in the best interest of the customer to actually go off and build those things or dedicate resources to those capabilities themselves or keep those resources dedicated all the time. And so you've got a couple of options. The first option is to leverage a public cloud type of technology, the IBM dev test cloud, Amazon EC2, or these types of transient workloads. So if you need to do a high scalability test on a thousand JVMs, rather than dedicate a thousand boxes to your application, you can create a thousand EC2 instances. And from a web server perspective, middleware services in the cloud is being one of the major pillars. We've got significant WebSphere support out on EC2. We just launched our IBM test dev cloud for a cloud just recently.

And so the number of things even from the app lifecycle process that become interesting for public cloud and cloud deploys. So let's say if a customer is to immediately take advantage of the cloud for those two topics to consider, evolving and incrementally enabling cloud functionality within the data center, and then better or streamlining the cost around application development so that they're able to move transient types of workloads out into the public cloud, for example, to reduce their time to value -- or improvement time to value reducing cost around testing ability and bringing their apps to production.

TV: Yeah, you mentioned disciplines and methodologies. I think those are going to be extremely critical in any sort of cloud adoption strategy. How might someone find additional information on the WebSphere cloud strategy and the overall components of IBM's offering?

SA: Sure. So one of the easiest things to do is just search IBM Cloud, two words, IBM and the word cloud and you'll get to the broader IBM cloud offerings. And if you Google or if you search for IBM WebSphere cloud, you'll get to -- there's a dedicated developer workspace that has all of our material for WebSphere specific topics around cloud. So you can start there and should be able to drill down and find other materials related to these topics.

TV: Thanks. That concludes this session of the Cloud QCamp Podcast Series. I want to thank my guest Snehal Antani and pilamaya yelo (thank you) for tuning into this particular session. Be sure to tune in for our next session in the Cloud QCamp Podcast Series. Thank you.

Announcer: This has been an ebizQ podcast.

Tim Vibbert

Tim Vibbert is the CTO of Red Eagle Service. Red Eagle Services is a leader for Service-Oriented Architecture (SOA)and Cloud Computing (CC) expertise and professional services. As a thought leader and authority on SOA and Cloud Computing, Red Eagle Services provides its clients strategic, adviory, and professional services.

Tim has over 15 years of experience in Enterprise IT in several industries including (Health Care, Legal, Manufacturing, Media Distribution, Government, and Retail). Twitter : @soachief

Tag Cloud

Leave a comment

Tim Vibbert

Tim Vibbert is the CTO of Red Eagle Service. Red Eagle Services is a leader for Service-Oriented Architecture (SOA)and Cloud Computing (CC) expertise and professional services. As a thought leader and authority on SOA and Cloud Computing, Red Eagle Services provides its clients strategic, adviory, and professional services.

Tim has over 15 years of experience in Enterprise IT in several industries including (Health Care, Legal, Manufacturing, Media Distribution, Government, and Retail). Twitter : @soachief

Recently Commented On

Monthly Archives