LifeFormulae Blog » Posts for tag 'libcloud'

What is Old Is New Again - The Cloud No comments yet

Cloud computing is the current IT rage, said to cure all information management skills.

Cloud computing is just a new name for timeshare, a system in which various entities shared a centralized computing facility. A giant piece or two of big iron and floors of tape decks provided information processing and storage capabilities for a price.

The user was connected to the mainframe by a dumb terminal and later on by PC’s. The advantage (said the sales jargon), was that the user didn’t need to buy any additional hardware, worry about software upgrades or data backup and recovery. They would only pay for the time and space their processes required. Resources would be pooled and connected by a high speed network and could be accessed as demanded. The user wouldn’t really know what computing resources were in use, they just got results. Everything depended on the network communications between the use and centralized computing source.

What is New

Cloud computing is more powerful today because the communications network is the Internet. Some Cloud platforms also offer Web access to the tools – programming language, database, web utilities needed to create the cloud application.

The most important aspect I believer the Cloud offers is instant elasticity. A process can be upgraded almost instantaneously to use more nodes and obtain more computing power.

There are quite a few blog entries out there concerning the “elastic” cloud. For thoughts on “spin up” and “spin down” elasticity see http://timothyfitz.wordpress.com/2009/02/14/cloud-elasticity/. For thoughts on “how elasticity could make you go broke, or On-demand IT overspending” see http://blogs.gartner.com/daryl_plummer/2009/03/11/cloud-elasticity-could-make-you-go-broke/.

And finally, an article that spawned the “elasticity is a myth” connotation or “over-subscriptionand over-capacity are two different things, see – http://www.rationalsurvivability.com/blog/?p=1672&cpage=1#comment-35881.

A good article that covers elasticity, hypervisors, and cloud security in general is located at http://queue.acm.org/detail.cfm?id=1794516. The queue.acm.org site is maintained by the Association for Computing Machinery. There are lots of articles on all sorts of computing topics including, “Why Cloud Computing Will Never Be Free” (http://queue.acm.org/detail.cfm?id=1772130).

The Clouds

The most notable Clouds are Amazon’s Elastic Cloud, Google’s App Engine, and Microsoft’s Azure.

The three Cloud delivery models include:

    • Software as a service (SaaS), applications running on a cloud are accessed via a web browser

    • Platform as a service (PaaS), cloud-developed user applications such as databases

    • Infrastructure as a service (IaaS), provides computing resources to users on an as-needed basis

Pros and Cons

There are pros and cons for Cloud Computing. Microsoft’s Bill Ballmer is a proponent of Cloud computing.

In a recent email (http://blog.seattlepi.com/microsoft/archives/196793.asp) to Microsoft’s employees, Ballmer make the following case for Cloud Computing. He advises his employees to watch a video (http://www.microsoft.com/presspass/presskits/cloud/videogallery.aspx) in which he makes the following points.

In my speech, I outlined the five dimensions that define the way people use and realize value in the cloud:

  • The cloud creates opportunities and responsibilities

  • The cloud learns and helps you learn, decide and take action

  • The cloud enhances your social and professional interactions

  • The cloud wants smarter devices

  • The cloud drives server advances that drive the cloud

Some very notable people are anti-cloud.

Richard Stallman, GNU software founder, said in recent interview for the London Guardian (http://www.guardian.co.uk/technology/2008/sep/29/cloud.computing.richard.stallman) that Cloud computing is a trap.

The Web-based programs like Google’s Gmail will force people to buy into locked, proprietary systems that will cost more and more over time, according to the free software campaigner.

‘It’s stupidity. It’s worse than stupidity: it’s a marketing hype campaign,’ he told The Guardian. ‘Somebody is saying this is inevitable — and whenever you hear somebody saying that, it’s very likely to be a set of businesses campaigning to make it true.’”

Aside from all that, what should a potential user be wary of in the Cloud? I’ll try to answer that below.

Security in the Cloud

Security in the cloud is a major concern. Hackers are salivating because everything – applications, data, are all in the same place.

How do you know the node your process is accessing is real or virtual? The Hypervisor (in Linux, a special version of the kernel) owns the hardware and spawns virtual nodes. If the Hypervisor is hacked, the hacker owns all the nodes created by it. http://www.linux-kvm. org has further explanations and discussions of virtual node creators/creations.

Data separation is a big concern. Could your data become contaminated by data in other environments in the cloud.? What access restrictions are in place to protect sensitive data?

Can a user in another cloud environment inadvertently or intentionally get access to your data?

Data interoperability is another question mark. A company cannot transfer data from a public cloud provider, such as Amazon, Microsoft, or Google, put it in a private IaasP that a private cloud provider develops for a company, and then copy that data from its private cloud to another cloud provider, public or private. This is difficult because there are no standards for operating in this hybrid environment.

Data Ownership

Who is the custodian and who controls data if your company uses cloud providers, public and private?

Ownership concerns have no been resolved by the cloud computing industry. At the same time, the industry has no idea when a standard will emerge to handle information exchanges.

W3C – http://www.w3.org/, is sponsoring workshops and publishing proposals concerning standards for the Cloud. You can subscribe to their weekly newsletter and stay up on all sorts of web-based technologies.

Also, the Distributed Management Task Force Inc.(http://www.dmtf.org/home) is a consortium ofof IT companies focusing on, “Developing management standards & promoting
interoperability for enterprise & Internet environments”.

The DMTF Open Cloud Standards Incubator was launched to address management interoperability for Cloud Systems (http://www.dmtf.org/about/cloud-incubator). The DMTF leadership board currently includes AMD, CA Technologies, Cisco, Citrix Systems, EMC, Fujitsu, HP, Hitachi, IBM, Intel, Microsoft, Novell, Rack Space, RedHat, Savvis, Sun Guard, Sun Microsystems, and VMWare.

Working with the Cloud

Working with the Cloud can be intimidating. One suggestion is to build a private cloud in-house before moving on to the public cloud.

However, even that has its difficulties. Not to worry, there are several tools available to ease the transition.

There is a Cloud programming language – Bloom, developed at UC Berkeley by Dr. Joseph Hellerstein. HPC In The Cloud has published an interview with Dr. Hellerstein at http://www.hpcinthecloud.com/features/Clouds-New-Language-Set-to-Bloom-92130384.html?viewAll=y

Bloom is based on Hadoop (http://hadoop.apache.org) which is open source software for High Performace Computing (HPC) from Apache..

For ease of inter connectivity, Apache has released Apache libcloud, a standard client library written in python for many popular cloud providers – http://incubator.apache.org/libcloud/index.html. But libcloud doesn’t cover data standards, just connectivity.

MIT StarCluster– http://web.mit.edu/stardev/cluster , is an open source utility for creating and managing general purpose computing clusters hosted on Amazon’s Elastic Compute Cloud (EC2). StarCluster minimizes the administrative overhead associated with obtaining, configuring, and managing a traditional computing cluster used in research labs or for general distributed computing applications.

All that’s needed to get started with your own personal computing cluster on EC2 is an Amazon AWS account and StarCluster.

HPC presents use cases as a means to understanding cloud computing. http://www.hpcinthecloud.com/features/25-Sources-for-In-Depth-HPC-Cloud-Use-Cases-93886489.html.

BCM Bioinformatics has a new methodology article – Cloud Computing for Comparative Genomics that includes a cost analysis of using the cloud. Download the .pdf at http://www.biomedcentral.com/1471-2105/11/259/abstract.

I hope this will get you started.  Once again, a big thanks to Bill for his assistance.

Top of page / Subscribe to new Entries (RSS)