Cloud Computing Simplified (to a degree)

Posted on Sep 8, 2009 in Records Management News

I found this to be some­what sim­ple depic­tion of this very inter­est­ing and grow­ing phenomenon.

CLOUD COMPUTING FROM WIKIPEDIA

Cloud com­put­ing is a style of com­put­ing in which dynam­i­cally scal­able and often vir­tu­al­ized resources are pro­vided as a ser­vice over the Inter­net. Users need not have knowl­edge of, exper­tise in, or con­trol over the tech­nol­ogy infra­struc­ture in the “cloud” that sup­ports them.

The con­cept gen­er­ally incor­po­rates com­bi­na­tions of the fol­low­ing: * infra­struc­ture as a ser­vice * plat­form as a ser­vice * soft­ware as a service

Other recent tech­nolo­gies that rely on the Inter­net to sat­isfy the com­put­ing needs of users. Cloud com­put­ing ser­vices often pro­vide com­mon busi­ness appli­ca­tions online that are accessed from a web browser, while the soft­ware and data are stored on the servers.

The term cloud is used as a metaphor for the Inter­net, based on how the Inter­net is depicted in com­puter net­work dia­grams and is an abstrac­tion for the com­plex infra­struc­ture it conceals.

Char­ac­ter­is­tics

Cloud com­put­ing cus­tomers do not gen­er­ally own the phys­i­cal infra­struc­ture serv­ing as host to the soft­ware plat­form in ques­tion. Instead, they avoid cap­i­tal expen­di­ture by rent­ing usage from a third-party provider. They con­sume resources as a ser­vice and pay only for resources that they use. Many cloud-computing offer­ings employ the util­ity com­put­ing model, which is anal­o­gous to how tra­di­tional util­ity ser­vices (such as elec­tric­ity) are con­sumed, while oth­ers bill on a sub­scrip­tion basis. Shar­ing “per­ish­able and intan­gi­ble” com­put­ing power among mul­ti­ple ten­ants can improve uti­liza­tion rates, as servers are not unnec­es­sar­ily left idle (which can reduce costs sig­nif­i­cantly while increas­ing the speed of appli­ca­tion devel­op­ment). A side effect of this approach is that over­all com­puter usage rises dra­mat­i­cally, as cus­tomers do not have to engi­neer for peak load limits.[15] Addi­tion­ally, “increased high-speed band­width” makes it pos­si­ble to receive the same response times from cen­tral­ized infra­struc­ture at other sites.

Eco­nom­ics

Cloud com­put­ing users can avoid cap­i­tal expen­di­ture (CapEx) on hard­ware, soft­ware, and ser­vices when they pay a provider only for what they use. Con­sump­tion is usu­ally billed on a util­ity (e.g. resources con­sumed, like elec­tric­ity) or sub­scrip­tion (e.g. time based, like a news­pa­per) basis with lit­tle or no upfront cost. A few cloud providers are now begin­ning to offer the ser­vice for a flat monthly fee as opposed to on a util­ity billing basis. Other ben­e­fits of this time shar­ing style approach are low bar­ri­ers to entry, shared infra­struc­ture and costs, low man­age­ment over­head, and imme­di­ate access to a broad range of appli­ca­tions. Users can gen­er­ally ter­mi­nate the con­tract at any time (thereby avoid­ing return on invest­ment risk and uncer­tainty) and the ser­vices are often cov­ered by ser­vice level agree­ments (SLAs) with finan­cial penalties.

Accord­ing to Nicholas Carr, the strate­gic impor­tance of infor­ma­tion tech­nol­ogy is dimin­ish­ing as it becomes stan­dard­ized and less expen­sive. He argues that the cloud com­put­ing par­a­digm shift is sim­i­lar to the dis­place­ment of elec­tric­ity gen­er­a­tors by elec­tric­ity grids early in the 20th century.[18]

Although com­pa­nies might be able to save on upfront cap­i­tal expen­di­tures, they might not save much and might actu­ally pay more for oper­at­ing expenses. In sit­u­a­tions where the cap­i­tal expense would be rel­a­tively small, or where the orga­ni­za­tion has more flex­i­bil­ity in their cap­i­tal bud­get than their oper­at­ing bud­get, the cloud model might not make great fis­cal sense. Other fac­tors impact­ing the scale of any poten­tial cost sav­ings include the effi­ciency of a company’s data cen­ter as com­pared to the cloud vendor’s, the company’s exist­ing oper­at­ing costs, the level of adop­tion of cloud com­put­ing, and the type of func­tion­al­ity being hosted in the cloud. [19][20]

Com­pa­nies

The “big four” of cloud com­put­ing ser­vices are said to be Ama­zon, Google, Microsoft and Salesforce.com. Cloud com­put­ing is also being adopted by indi­vid­ual users through large enter­prise cus­tomers includ­ing Gen­eral Elec­tric, Proc­ter & Gam­ble and Valeo .

Archi­tec­ture

The major­ity of cloud com­put­ing infra­struc­ture, as of 2000 con­sists of reli­able ser­vices deliv­ered through data cen­ters and built on servers with dif­fer­ent lev­els of vir­tu­al­iza­tion tech­nolo­gies. The ser­vices are acces­si­ble any­where that pro­vides access to net­work­ing infra­struc­ture. Clouds often appear as sin­gle points of access for all con­sumers’ com­put­ing needs. Com­mer­cial offer­ings are gen­er­ally expected to meet qual­ity of ser­vice (QoS) require­ments of cus­tomers and typ­i­cally offer SLAs.[25] Open stan­dards are crit­i­cal to the growth of cloud com­put­ing, and open source soft­ware has pro­vided the foun­da­tion for many cloud com­put­ing implementations.

His­tory

The Cloud is a term that bor­rows from tele­phony. Up to the 1990s, data cir­cuits (includ­ing those that car­ried Inter­net traf­fic) were hard-wired between des­ti­na­tions. Sub­se­quently, long-haul tele­phone com­pa­nies began offer­ing Vir­tual Pri­vate Net­work (VPN) ser­vice for data com­mu­ni­ca­tions. Tele­phone com­pa­nies were able to offer VPN based ser­vices with the same guar­an­teed band­width as fixed cir­cuits at a lower cost because they could switch traf­fic to bal­ance uti­liza­tion as they saw fit, thus uti­liz­ing their over­all net­work band­width more effec­tively. As a result of this arrange­ment, it was impos­si­ble to deter­mine in advance pre­cisely which paths the traf­fic would be routed over. The term “tele­com cloud” was used to describe this type of net­work­ing, and cloud com­put­ing is con­cep­tu­ally some­what similar.

Cloud com­put­ing relies heav­ily on vir­tual machines (VMs), which are spawned on demand to meet user needs. A com­mon depic­tion in net­work dia­grams is a cloud outline.

The under­ly­ing con­cept of cloud com­put­ing dates back to 1960, when John McCarthy opined that “com­pu­ta­tion may some­day be orga­nized as a pub­lic util­ity”; indeed it shares char­ac­ter­is­tics with ser­vice bureaus that date back to the 1960s. The term cloud had already come into com­mer­cial use in the early 1990s to refer to large Asyn­chro­nous Trans­fer Mode (ATM) net­works. Ill-fated startup Gen­eral Magic launched a short-lived cloud com­put­ing prod­uct in 1995 in part­ner­ship with sev­eral telecom­mu­ni­ca­tions com­pany part­ners such as AT&T, just before the consumer-oriented Inter­net became pop­u­lar. By the turn of the 21st cen­tury, the term “cloud com­put­ing” began to appear more widely,[28] although most of the focus at that time was lim­ited to SaaS.

In 1999, Salesforce.com was estab­lished by Marc Benioff, Parker Har­ris, and their asso­ciates. They applied many tech­nolo­gies devel­oped by com­pa­nies such as Google and Yahoo! to busi­ness appli­ca­tions. They also pro­vided the con­cept of “On demand” and SaaS with their real busi­ness and suc­cess­ful cus­tomers. The key for SaaS is that it is cus­tomiz­able by cus­tomers with lim­ited tech­ni­cal sup­port required. Busi­ness users have enthu­si­as­ti­cally wel­comed the result­ing flex­i­bil­ity and speed.

In the early 2000s, Microsoft extended the con­cept of SaaS through the devel­op­ment of web ser­vices. IBM detailed these con­cepts in 2001 in the Auto­nomic Com­put­ing Man­i­festo, which described advanced automa­tion tech­niques such as self-monitoring, self-healing, self-configuring, and self-optimizing in the man­age­ment of com­plex IT sys­tems with het­ero­ge­neous stor­age, servers, appli­ca­tions, net­works, secu­rity mech­a­nisms, and other sys­tem ele­ments that can be vir­tu­al­ized across an enterprise.

Ama­zon played a key role in the devel­op­ment of cloud com­put­ing by mod­ern­iz­ing their data cen­ters after the dot-com bub­ble which, like most com­puter net­works, were using as lit­tle as 10% of their capac­ity at any one time just to leave room for occa­sional spikes. Hav­ing found that the new cloud archi­tec­ture resulted in sig­nif­i­cant inter­nal effi­ciency improve­ments whereby, small, fast-moving “two-pizza teams” could add new fea­tures faster and eas­ier, Ama­zon started pro­vid­ing access to their sys­tems through Ama­zon Web Ser­vices on a util­ity com­put­ing basis in 2005.

In 2007, Google, IBM, and a num­ber of uni­ver­si­ties embarked on a large scale cloud com­put­ing research project,[30] around the time the term started, it was a hot topic. By mid-2008, cloud com­put­ing gained pop­u­lar­ity in the main­stream press, and numer­ous related events took place. ng will result in dra­matic growth in IT prod­ucts in some areas and in sig­nif­i­cant reduc­tions in other areas.”

In 2009, Cloud Com­put­ing Solu­tions by Google, Ama­zon, Microsoft, and IBM are the most pop­u­lar among users with Sun and Ubuntu fol­low­ing them in the Cloud.

Com­pli­ments of File­Man Rsearch … To Sub­scribe to the File­Man Blog click here … http://www.carymcgovern.com/feed/

Best Per­sonal Regards,

Cary