Adarsh Patil Grid Cloud On-Demand Computing

Land to Cloud

"Cloud is a collection and interconnection of utilities (mail, storage, compute power, databases, social networks, banking, software, applications, infrastructure, other commodity services and future commoditize'ble services) that follows and serves the human being on subscription" - Adarsh Patil

'Cloud is expensive for customers who "like" it and useful for customers who "need" it on demand' -Adarsh Patil (click here)

Cloud Computing Questions click here :)

  1. One powerful server doing the work for the dumb terminals : Clients assigning the work to the Server
  2. Many powerful clients doing work on behalf of the Server: Server assign the work to all the Clients
  3. Group of machines (loosely / tightly coupled) acting as one Power machine and serving all the clients via web or Desktop client
  4. Group of Data Centers hosting (un)limited Systems for computational storage, computational power, huge bandwidth and provisioning software to perform 1, 2, 3 and multi-tenancy ( having more than one application or operating system to run on the same system to improve resource utilization and ROI)

Multi-tenancy

Multi-tenancy is not new, it is age old. Now with every company from startups to large corporations are on to Cloud bandwagon, multi-tenancy is being made as a 'BIG BUBBLE'. Earlier, web hosting companies used to give an option to host a website on a shared server or a dedicated server. Share server hosting meaning multiple websites are hosted along with your website on one server. On the other hand dedicated server would host only your website on that server, it is expensive than shared hosting. Now this concept has extended to share multiple operating system, applications(a blade in a blade server rack of a Data center hosting thousands of blades) on one shared server which hosts virtual machines of yours as well as others. This multi-tenancy concept has been exploded into many concepts like PAAS, IAAS, SAAS which are explained below.

Cloud Popcorn

Present Cloud computing model is oligopoly (what is oligopoly?
) There are small number of (Cloud/Grid/CoGs/CoWs)Resource sellers or Infrastructure providers, so they can fix their own price until the ration of sellers to end users changes.
Reference link: GridEcon
Types of Cloud
  • Public Cloud(Virtual machine Instances)--Grid(Physical machines)
  • Private Cloud--Cluster(Grid - Virtual Organistaion)
  • Hybrid Cloud--Cluster of Clusters Idle processor harvesting or CPU cycle stealing (SETI@home)
  • Name your cloud--for example: Green Cloud, Community Cloud, Education Cloud, Entertainment Cloud
  • Cloud is kind of enemy of the hardware vendors, which does not allow them to potentially sell hardware.
  • Hardware is getting cheaper day by day and numbers of cores on the Desktop machine are increasing, (Intel 48Core, Intel i9/i7/i5,AMD) until one day everyone (every home) has a powerful computer equivalent to the current trend supercomputer or cloud computer.
  • Then hardware will become the enemy of Software (managing and running the cloud) eventually.
  • What has been said here, is just mapping the consequenses of increasing cheaper availability of powerful hardware at disposal and EaaS model ( Everything as a Servic)

Cloud: The way we see, architect and use technology of the past, present and future

a) Offloading or Outsourcing local work onto a cloud (data centers) is Cloud Computing. Reason might be local work needs more computing power, storage space or needs an elastic or on demand option to expand its resource capability.

b) Running multiple instances of operating system(s)(might be same/different - multi-tenancy) on the same physical hardware for doing STEP (a) is called Cloud Computing.

c) Extending ones capability in terms of storage, computational needs, availability, seamless access, scalability Step (a) and Step (b) helps.

d) SaaS, PaaS, IaaS are the characteristics of the cloud.

e) Resource utilization(for ROI) in Cluster / Grid computing was through harvesting of idle processor cycles via client, screensaver, conscription based for doing computations. In Cloud computing resource utilization is through running more than virtual operating systems (Windows / Linux / Mac OSX) on the same physical hardware. example: AMI (Amazon EC2).

f) Cloud is elastic and unlimited (increase and decrease resources at your finger tips) where as cluster/Grid resources are not elastic and limited.

e)
  • with respect to the Cloud, the workflow of getting 1000 machines(images) or cores up and running are
    • login and give your credit card details and sign up
    • via their api or tools or plug-in launch 1000 machines/cores, they will be online in 15 to 18 minutes
    • do your computations and shut the machines down
    • get your bill
  • with respect to Grid, the workflow of getting 1000 machines(real physical) or cores up and running are
    • Get your certificates from the CA manager
    • Send the CA your passport or drivers license
    • wait for a week or a month(s)
    • you may or may not get access to those machines
    • after passing a list of hurdles of who you are is who you are
    • you will get access
    • I am not sure whether you will be billed thats a plus point here
SaaS - Software as a Service
PaaS - Platform as a Service
IaaS - Infrastructure as a Service (may be Integration as a Service)

Cloud(IaaS in mind) is kind of a super mall grocery store with kitchen(for example) where one can get everything(related to cooking for example)and a platform(kitchen) to develop(in case of software/OS/application)/cook(for example) your own applications/cuisine and use their infrastructure. They also give you fire extinguisher/security for your environment(for example your kitchen). The only thing is you have to trust all the commodities available from their grocery store, you have to trust that all of your recipes are safe/safely stored in that environment. If the cloud is lost , you are lost.
Kind of liked the below picture. Image Reference:http://blog.gogrid.com/cloud-pyramid/

Nice visual divisions of the cloud: Credits to BEA Systems and Authors: Peter Laird, Kent Dickson

Green Computing/ Green Grid / Recession based Computing™

The characteristic of a software that is able to work with existing hardware/software/application infrastructure and does not need any additional specialized hardware/software as a dependency and is able make profitable return of interest(ROI) is termed as Green Computing. This kind of software enables Green Grid. Green in the sense, nothing is consumed from the environment but something under-utilized is being used. There are many definitions or variety of definitions, use the one which you are comfortable with. There might be another "terminology" on the way ( Distributed Computing --> Grid --> Cloud --> Green --> what next? )

Cloud(earlier Grid) Concept

Grid Computing Concept by Adarsh

What is Grid?

Grid is a shared collection of reliable (cluster-tightly coupled) & unreliable resources (loosely coupled machines) and interactively communicating researchers of different virtual organisations (doctors, biologists, physicists). Grid System controls and coordinates the integrity of the Grid by balancing the usage of reliable and unreliable resources among its participants providing better quality of service. It means Grid Systems are synonymous to operating systems that control, co-ordinates and schedules the resources of a system.

"In 1854, George Boole (1815-1864), Professor of Mathematics at Cork from 1849 despite having no first degree, formalised a set of such rules in the seminal work entitled, perhaps optimistically, An Investigation of the Laws of Thought. Boole's aim was to identify the rules of reasoning in a rigorous framework and revolutionised formal logic after thousands of years of little progress. They transformed logic from a philosophical into a mathematical discipline. These rules have subsequently become known as Boolean algebra and the design of all modern binary digital computers has depended on the results of this work. These logical operations, normally implemented as electronic gates, are all that are required to perform more complicated operations such as arithmetic." The Virtual Museum of Computing.

I'm Adarsh Patil,  was a research scholar at University College Cork. I was doing PhD under the supervision of Dr. John Morrison in the group Centre for Unified computing. I started my PhD in January 2002. and currently waiting for thesis correction and defense. I have chosen Grid computing and Cluster Computing as my research area. Most of the work in our research group is based on Condensed Graph Model of Computing, which is the work of our supervisor's PhD. Dr.John Morrison (my supervisor) was supervised by one of the three postgraduate students of Edsger W. Dijkstra.

Centre for Unified Computing (CUC) is one of the core grid and cluster computing research group in Ireland whose head and director is Dr. John Morrison. Its a part of Grid-Ireland. Other core research groups of Grid-Ireland are Grid Group from Trinity College Dublin (TCD) headed by Dr. Brian Coghlan, Grid Application group from National University of Ireland Galway(NUIG) headed by Dr. Andrew Shearer and Department of Computer Science, Queens University Belfast(Northern Ireland) headed by Prof.Ron Perrott.

Our core research work is on WebCom, whose chief architects are Dr. John Morrison , Dr David Power , Dr. James Kennedy , Dr. Keith Power and other researchers (Adarsh, Padraig, Brian, Sunil, Philip, Neil, John O' Regan, Tom, Barry, Max and Hongbin) of the group are currently building their research work around it. There is a security team in this research group headed by Dr. Simon Foley. Their security models are based on condensed graph model. We are all currently working with other research groups across Ireland on a project called WebCom-G.

Connect and ContactView Adarsh Patil (adarshvp@gmail.com)'s profile on LinkedIn

Buzzzzz!

Cloud Computing is backed up by the Industry whereas Grid Computing is backed up by the academic institutions. They are both coming from different point of views but they might converge or remain as they are?.

News

"Grid can be termed as a platform to support hardware and software virtualization"




Grid Computing Questions | Grid Computing Books | Grid Links | Contact

Thank you for visiting my website
Site Design by Adarsh Patil
Copyrights reserved © CUC & Adarsh