Posts tagged: cloud architecture

Cloud Architecture Design: Should it be Top-Down or Bottom-Up?

By Steve Jin, July 26, 2010

In my last blog, I discussed how to optimize workloads across the cloud. This is based on the assumption that you already have an existing infrastructure. What if you don’t have an existing cloud infrastructure but would like to design one from scratch? Here is what you should be thinking about to get the most from your new cloud.

But first let’s take a look at other types of infrastructures – say a road. When you design a new road, you have to collect data such as population densities around the area, people’s working schedules, what types of vehicles will run on the road, and so on. With that information, you can decide how many lanes you want, what kind of road surface is required, and so on. You don’t just make up the design specification from scratch, and lay down an eight-lane freeway everywhere.

The same process applies in designing the cloud infrastructure as well. Unfortunately this is not what we see often today.

Top-down approach

In my previous blog , I said infrastructure is a means and application is the end. We need to drive the design cloud architecture from the application perspective. This is what I call the top-down approach. Read more »

Author: Steve Jin is the author of VMware VI and vSphere SDK (Prentice Hall), creator of VMware vSphere Java API. For future articles, please subscribe to Email or RSS, and follow on Twitter.

Decomposition and Challenges in Parallel Programming: Is It Useful for Cloud Computing?

By Steve Jin, June 23, 2010

A recent article from Dr. Dobb’s introduced Fundamental Concepts of Parallel Programming. Richard Gerber and Andrew Binstock, authors of Programming with Hyper-Threading Technology, discussed three different forms of de-compositions for multi-threading:

  1. Functional decomposition. It’s one of the most common ways to achieve parallel execution. Using this approach, individual tasks are catalogued. If two of them can run concurrently, they are scheduled to do so by the developer.
  2. Producer/Consumer. It’s a form of functional decomposition in which one thread’s output is the input to a second. Can be hard to avoid, but frequently detrimental to performance.
  3. Data decomposition, a.k.a. “data level parallelism.” It breaks down tasks by the data they work on, rather than by nature of the task. Programs that are broken down via data decomposition generally have many threads performing the same work, just on different data items.

To make the three forms easy to understand, the authors used gardening as analogy where the threads map to gardners. For exmaple, the fuctional decomposition in gardening is to have one gardner to move the lawn and the other to weed. I find this analogy very intuitive and easy to follow. Even you don’t know multh-threading, you can guess it out from the gardening analogy.

The challenges while working with multi-threading are: Read more »

Author: Steve Jin is the author of VMware VI and vSphere SDK (Prentice Hall), creator of VMware vSphere Java API. For future articles, please subscribe to Email or RSS, and follow on Twitter.

Building Do-It-Yourself PaaS: My VMworld Session Proposals

By Steve Jin, May 19, 2010

Most people who are interested in VMworld already know the public voting for the proposal is now open till 26th. If you would like to hear about specific topics, it’s high time to cast your votes.

For each track, all the presentation proposals are listed together in one page. To quickly locate a particular proposal, you can use find feature of your browser. Once you login, I would suggest to browse all the proposals and vote for those you find useful. Casting a vote is just two mouse clicks: one for voting and the other to close the confirmation message box.

Two of my proposals are there in two different tracks: public cloud and enterprise application. Please feel free to vote for them.

Title: Building Do-It-Yourself PaaS Platform

Session Id: PC7724

Abstract:

“More and more modern applications are built on top of various middleware for productivity and portability. With cloud computing, the service providers host these middleware and provide them as PaaS (Platform As A Service) services to application developers. PaaS has many benefits, for example no more investment on datacenter and operation, no worry about scalability. The disadvantage is also obvious ñ you are locked in with the PaaS vendor. This session presents a new way, DIY PaaS, to build your applications. It allows you to control the middleware and package them into virtual machines which be provisioned both at IaaS public cloud or inside enterprise. You will have a little more management work in so doing than using typical PaaS service, but you wonít be locked with any vendor with full control over your stack and flexibility to move your applications to and out the public cloud. A good DIY PaaS framework can mitigate the extra work to the minimum and automate the cloud bursting for the best of the application owners. The speaker will guide you through best practices in architecting and building the DIY PaaS and building applications on top of it. At the end, you see a demo showing how it can be done with a real project.”

Track: Hybrid and Public Cloud

http://www.vmworld.com/community/conferences/2010/cfpvote/hapcloud

Title: Re-architecting Enterprise Applications for vSphere

Session Id: EA7723

Abstract:

“While legacy enterprise applications continue to run on VMware vSphere, they do not fully leverage the benefits of virtualization. For the best performance and scalability, enterprise application should be architected differently to run on vSphere. This represents new challenges and opportunities for enterprise application development. This session discusses in detail what these challenges and opportunities are. It introduces top 10 proven best practices in architecting enterprise applications in light of vSphere, and more importantly how to apply these best practices in real projects. In the end, it presents an architectural blueprint illustrating how different best practices work together for best enterprise applications running on vSphere. Top 10 Best Practices include: 1. Move Up to Higher Level Software Stack with Middleware and Framework 2. Donít Assume Anything 3. De-couple Your Application 4. Scale Application As Needed 5. Build With Failure In Mind 6. Secure Your Application 7. Leverage vApp 8. Streamline the Packaging & Deployment Cycle 9. Manage Your Applications 10. Keep It Simple”

Track: Enterprise Applications

http://www.vmworld.com/community/conferences/2010/cfpvote/eapplications

Author: Steve Jin is the author of VMware VI and vSphere SDK (Prentice Hall), creator of VMware vSphere Java API. For future articles, please subscribe to Email or RSS, and follow on Twitter.

Cloud Computing: As Service or As Architecture?

By Steve Jin, April 11, 2010

Last week I saw an incoming link at my blog site from privatecloud.com, a website promoting private cloud computing inside enterprises backed by EMC. Due to curiosity, I browsed the website and found a video by VMware CEO Paul Maritz on cloud computing. BTW, my website is also featured at the home page. Thanks privatecloud.com!

In the video, Paul talked about the cloud computing, mostly referring to services over the Internet, can also be an architecture pattern for enterprises. When that architecture is implemented, you will have a private cloud on premise. Although not using cloud services from any service provider, you still get almost all the benefits in a private cloud.

From the system architecture perspective, your applications built within a private cloud aren’t much different than those built using external cloud services. If both of the services follow the same interface spec, your application may switch between public cloud and private cloud either statically or dynamically. Read more »

Author: Steve Jin is the author of VMware VI and vSphere SDK (Prentice Hall), creator of VMware vSphere Java API. For future articles, please subscribe to Email or RSS, and follow on Twitter.

VMware PEX 2010 – Day Three

By Steve Jin, February 12, 2010

I know I haven’t written the day three of my PEX yet. You know there was a celebration party at that night and we had a really good time there. Now I’m back and have more time to write about day three.

CTO Steve Herrod’s Keynote

This is the most important part of day three, followed by the celebration party:-). If you are interested in knowing what is happening in VMware R&D and what new products are coming from VMware, you don’t want to miss a single minute of Steve’s keynote. That is why I got there early and sat in the front. Read more »

Author: Steve Jin is the author of VMware VI and vSphere SDK (Prentice Hall), creator of VMware vSphere Java API. For future articles, please subscribe to Email or RSS, and follow on Twitter.

Page 1 of 11

OfficeFolders theme by Themocracy