Wišniewski, Michał
A icle
Cloud compu ing as a ool o imp o ing business
compe i i eness
Founda ions o Managemen
P o ided in Coope a ion wi h:
Facul y o Managemen , Wa saw Uni e si y o Technology
Sugges ed Ci a ion: Wišniewski, Michał (2013) : Cloud compu ing as a ool o imp o ing business
compe i i eness, Founda ions o Managemen , ISSN 2300-5661, De G uy e , Wa saw, Vol. 5, Iss. 3,
pp. 75-88,
h ps://doi.o g/10.2478/ man-2014-0021
This Ve sion is a ailable a :
h ps://hdl.handle.ne /10419/184563
S anda d-Nu zungsbedingungen:
Die Dokumen e au EconS o dü en zu eigenen wissenscha lichen
Zwecken und zum P i a geb auch gespeiche und kopie we den.
Sie dü en die Dokumen e nich ü ö en liche ode komme zielle
Zwecke e iel äl igen, ö en lich auss ellen, ö en lich zugänglich
machen, e eiben ode ande wei ig nu zen.
So e n die Ve asse die Dokumen e un e Open-Con en -Lizenzen
(insbesonde e CC-Lizenzen) zu Ve ügung ges ell haben soll en,
gel en abweichend on diesen Nu zungsbedingungen die in de do
genann en Lizenz gewäh en Nu zungs ech e.
Te ms o use:
Documen s in EconS o may be sa ed and copied o you pe sonal
and schola ly pu poses.
You a e no o copy documen s o public o comme cial pu poses, o
exhibi he documen s publicly, o make hem publicly a ailable on he
in e ne , o o dis ibu e o o he wise use he documen s in public.
I he documen s ha e been made a ailable unde an Open Con en
Licence (especially C ea i e Commons Licences), you may exe cise
u he usage igh s as speci ied in he indica ed licence.
h ps://c ea i ecommons.o g/licenses/by-nc-nd/3.0
Founda ions o Managemen , Vol. 5, No. 3 (2013), ISSN 2080-7279
DOI: 10.2478/ man-2014-0021 75
CLOUD COMPUTING AS A TOOL FOR IMPROVING BUSINESS COMPETITIVENESS
Michał WIŚNIEWSKI
Facul y o Managemen , Wa saw Uni e si y o Technology, Wa saw, Poland
e-mail: m[email p o ec ed].edu.pl
Abs ac : This a icle o ganizes knowledge on cloud compu ing p esen ing he classi ica ion o deploymen
models, cha ac e is ics and se ice models. The au ho , looking a he p oblem om he en ep eneu ’s pe -
spec i e, d aws a en ion o he di e ences in he bene i s depending on he cloud compu ing deploymen
models and conside s an e ec i e way o selec ion o cloud compu ing se ices acco ding o he speci ici y
o o ganiza ion. Wi hin his wo k, a hesis s a emen was conside ed ha in economic e ms he cloud com-
pu ing is no always he bes solu ion o you o ganiza ion. This aises he ques ion, “Wha kind o ools
should be used o es ima e he use ulness o he model cloud compu ing se ices in he en e p ise?”
Keywo ds: cloud compu ing, da a p ocessing, oppo uni ies, h ea s, se ice models, deploymen models, he
cha ac e is ics o cloud compu ing, inno a ion, es ima ing ope a ing cos s.
1 In oduc ion
One o he mos impo an esou ces in he mode n
economy is in o ma ion. O ganiza ions depend on how
as ly he da a a e accessed as he inished p oduc ,
which is use ul o decision making a all le els
o managemen : s a egic, ac ical and ope a ional le el.
A he same ime, one o he p oblems o any o ganiza-
ion is an excess o da a. To e ec i ely use he da a,
hey mus be collec ed, s o ed and p ocessed wi h he
suppo o ools such as: da abases, da a wa ehouses
and BI sys ems. These ools allow as access o da a
and p esen hem in a anspa en manne , which acili-
a es he wo k o , among o he s, manage s.
Such a me hod o da a p ocessing equi es an app op i-
a e in as uc u e. To begin wi h, special ooms de-
signed o se e ooms o he se e s, disk a ays,
backup copy sys ems, ending wi h he clien s a ions
should be main ained. All elemen s o he ne wo k
gene a e cos s o he o ganiza ion a he ime o pu -
chase, du ing use and a a ime when no longe use ul.
So wa e licenses also inc ease cos s associa ed wi h
he IT sec o . Some o ganiza ions, especially small, a e
no able o mee hese cos s losing he oppo uni y
o e ec i e compe i ion in he ma ke place.
A emedy o he high cos s o collec ing and p o-
cessing da a wi hin he o ganiza ion may be o build
i ual se e s in he en i onmen o cloud compu ing.
The e a e companies ha o e he oppo uni y o make
use o ha dwa e esou ces, ope a ing pla o ms
and so wa e as a paid se ice ia he In e ne .
The ecipien is obliged o co e only he cos s o pu -
chase and main enance o e minal s a ions, he In e -
ne , cloud compu ing se ice cos s and main aining
a minimum IT le el dependen on he size and needs
o he o ganiza ion.
Looking a he phenomenon o cloud compu ing om
he en ep eneu ’s pe spec i e pu chasing cloud com-
pu ing se ice, a ques ion a ises whe he his se ice
should eplace he collec ion and p ocessing o da a
wi hin he o ganiza ion? Wha a e he oppo uni ies
and isks, bene i s and obliga ions coming ou o he
change o da a p ocessing?
This a icle p esen s he cu en s a e o knowledge
on he concep o cloud compu ing and iden i ies ques-
ions ha companies need o answe conside ing how
o p ocess he da a wi hin i s own s uc u e. Con a y
o popula opinion, p esen ed in li e a u e and du ing
con e ence speeches, he au ho belie es ha he cu -
en le el o de elopmen o he cloud compu ing se -
ice canno , in any case, be an al e na i e o he
exis ing cen e s in he o ganiza ions. F om he poin
o iew o he en ep eneu , he mos impo an indica-
o in choosing he me hod o da a p ocessing is
he le el o cos s. Poo ly chosen cloud compu ing se -
ice may esul in inc eased cos s, which nega i ely
a ec he compe i i eness o he company.
2 Cloud Compu ing
2.1 De ini ions
Cloud compu ing idea is no new. The concep o cloud
compu ing is ha IT esou ces, bo h ha dwa e
and so wa e, a e ea ed by he use as ano he medium
such as elec ici y, gas o wa e . The ecipien does no
physically wonde s whe e elec ici y comes om in his
76 Michał Wiśniewski
socke , jus pays o he se ice. Mos o us use cloud
compu ing wi hou being awa e o i . A pe ec exam-
ple is he cloud compu ing e-mail se ice. The majo i y
o use s do no know whe e physically loca ed e-mail
messages a e. Using he mailbox is done o e he In e -
ne , ega dless o he loca ion o de ice you a e using;
you always ha e access o e-mails. The e o e, he mail
sys em supplie p o ides o us a se ice. The e y
concep o cloud compu ing has been used o he i s
ime by Ne Cen icin 1997, so i is ela i ely new
and s ill e ol ing, making i di icul o ind one ecog-
nized by all de ini ion in he li e a u e.
Acco ding o IBM, cloud compu ing is a new model
o IT use and p ocessing s yle in which he business
p ocesses, applica ions, da a and IT esou ces a e de-
li e ed o use s in he o m o se ices [5, p. 145]. An-
o he de ini ion comes om he po al Wikipedia:
Cloud compu ing is a p ocessing model based on he
use o se ices p o ided by ex e nal o ganiza ions.
Func ionali y is unde s ood as a se ice (which gi es
added alue o he use ) o e ed by he so wa e (and
he necessa y in as uc u e) [15]. The accep ed de ini-
ion o cloud compu ing in he scien i ic s udies is ha
one de eloped by he Na ional Ins i u e o S anda ds
and Technology (NIST) in 2011. Acco ding o he
NIST, cloud compu ing is a model o enabling ubiqui-
ous, con enien , on-demand ne wo k access o
a sha ed pool o con igu able compu ing esou ces (e.g.
ne wo ks, se e s, s o age, applica ions and se ices)
ha can be apidly p o isioned and eleased wi h min-
imal managemen e o o se ice p o ide in e ac ion.
This cloud model is composed o i e essen ial cha ac-
e is ics, h ee se ice models and ou deploymen
models [10, p. 3].
All de ini ions consequen ly main ain ha he objec
se ices a e b oadly de ined IT esou ces. Howe e ,
only he NIST de ini ion emphasizes he ac o ha
dis inguishes cloud compu ing om o he o ms
o ou sou cing. I e e s o he s a emen ha he se ice
is a ailable on-demand, wi hou necessa y ea lie o -
de s o s a ing ee as i ake place, o example, in he
co-loca ion model. Ano he impo an ea u e is he
me hod o calcula ing ees. In con as o classic o ms
o IT ou sou cing, a cloud compu ing model p o ides
a di e en solu ion whe e we pay o he esou ces
ac ually used, and no o hei decla ed use.
Looking a cloud compu ing om a pu ely echnical
poin o iew, his se ice is a blend o he known solu-
ions in he p o ision o IT se ices, o example:
U ili y compu ing has usually en isioned some o m
o i ualiza ion so ha he amoun o s o age
o compu ing powe is a ailable o indi idual busi-
ness uni s o p ocedu es. I also allows ho ough se -
lemen o he amoun o p ocessed da a [13].
G id compu ing is de ined as a ha dwa e and so -
wa e in as uc u e ha p o ides dependable, con-
sis en , easily cus omizable and a o dable access
o compu ing esou ces [1, pp. 749 – 771]. In con-
as o u ili y, compu ing can also use he compu-
ing powe o PC.
Ou sou cing is he ans e o ex e nal se ice p o-
ide s in acco dance wi h he p o isions in he
ag eemen (con ac ) epea ed in e nal asks o an
o ganiza ion ela ed o he implemen a ion o s a ,
equipmen , acili ies, equipmen , echnology
and o he esou ces and decision-making powe s e-
ga ding hei use [3, p. 3].
Vi ualiza ion is cha ac e ized by abs ac pe cep-
ion o ha dwa e esou ces so ha he esou ce can
be ac i a ed o ope a e in a ious ways, ega dless
o i s physical o m o loca ion [9, p. 856]. This al-
lows mo e e icien use o exis ing ha dwa e en i-
onmen esou ces by modi ying he cha ac e is ics
o any esou ces, adap ing hem o use equi e-
men s.
Dis ibu ed compu ing e m conceals he echnique
o unning calcula ions, enabling he sha ing
o compu ing esou ces such as: p i a e compu e s,
se e s, e c., which is usually geog aphically dis-
pe sed. In a gene al sense, his echnology means
he da a p ocessing ega ded as a public se ice.
In o he wo ds, he clien does no ma e whe e
he da a a e s o ed o which compu e does he job.
The cus ome can o de he in o ma ion o he cal-
cula ions and esul s in he scope and ime equi ed
[4, p. 37].
The unc ionali y o hese concep s allows o inno a-
ion wi hin he model o business managemen .
The a ailabili y o IT esou ces on demand and bea ing
he cha ges o he ac ually used esou ces allows you
o change he cos s uc u e o he o ganiza ion. I leads
o he educ ion o he g oup cos classi ied as capi al
expendi u es (CAPEX)1 and inc easing o ope a ing
1 CAPEX- expendi u es alloca ed o buy ixed asse s such as
equipmen , e.g. IT equipmen .
Cloud Compu ing as a Tool o Imp o ing Business Compe i i eness 77
Figu e 1. Concep o cloud compu ing se ices
(sou ce: own elabo a ion)
expendi u es (OPEX).2 This means ha he ba ie
ela ed o he pu chase o IT in as uc u e o he neces-
sa y high cos s o he en in ad ance ceases o exis .
A company using he cloud compu ing model bea s he
cha ges on an ongoing basis, as pa o he ope a ing
cos s.
The basic p emise o he cloud compu ing is he a aila-
bili y o se ices deli e ed o e he In e ne o in ane ,
based on a model composed o i e essen ial cha ac e -
is ics, h ee se ice models and ou deploymen mod-
els shown in Figu e 1.
O ganiza ions aking he decision o use cloud compu-
ing se ices do no e lec ha kind o cloud is o le
sui able. This leads o ine icien use o IT, and in ex-
eme cases, o inc ease in he ope a ing cos s o hose
depa men s. To a oid such si ua ions, we mus unde -
s and he ad an ages and disad an ages o he di e en
ypes o clouds. Fo his pu pose, he elemen s shown
in Figu e 1 will be discussed in mo e de ail.
2.2 Essen ial Cha ac e is ics
On-demand sel -se ice - a consume can unila e al-
ly con igu e p o ision compu ing capabili ies, such
as se e ime and ne wo k s o age, as needed au-
2 OPEX – is he sum o a business's ope a ing expenses o
a pe iod o ime.
oma ically wi hou equi ing human in e ac ion
wi h each se ice p o ide .
B oad ne wo k access - capabili ies a e a ailable
o e he ne wo k and accessed h ough s anda d
mechanisms ha p omo e use by he e ogeneous
hin3 o hick clien 4 pla o ms (e.g. mobile phones,
able s, lap ops and wo ks a ions).
Resou ce pooling - he p o ide ’s compu ing e-
sou ces a e pooled o se e mul iple consume s us-
ing a mul i- enan model, wi h di e en physical
and i ual esou ces dynamically assigned and e-
assigned acco ding o consume demand.
Rapid elas ici y - capabili ies can be elas ically p o-
isioned and eleased, in some cases au oma ically,
o scale apidly ou wa d and inwa d commensu a e
wi h demand.
3 Thin clien is a compu e o specialized de ice (compu e
e minal) and ela ed so wa e clien , enabling suppo o appli-
ca ions c ea ed in he clien –se e a chi ec u e. A hin clien
occu s as independen o he suppo ed applica ion se e ( he
change does no in ol e ha ing o change clien so wa e). An-
o he ad an age is he low demand o p ocessing powe [15].
4 Thick clien is a compu e in clien –se e a chi ec u e o ne-
wo ks. I e e s o he use 's wo ks a ion, which is equipped wi h
a se o pe iphe als, ope a ing sys em ins alled and se o applica-
ions. P og ams a e execu ed di ec ly and au onomously a he
s a ion does da a p ocessing and da a exchange wi h he use and
o he compu e s on he ne wo k. Da a a e s o ed on he se e
side. [15]
78 Michał Wiśniewski
Measu ed se ice - cloud sys ems au oma ically con ol
and op imize esou ce use by le e aging a me e ing
capabili y a some le el o abs ac ion app op ia e
o he ype o se ice (e.g. s o age, p ocessing, band-
wid h, compu ing powe , RAM and ac i e use ac-
coun s). I p o ides anspa ency o bo h he p o ide
and consume because o he pay-pe -use s uc u e.
The au ho e e s o he de ini ions o he cha ac e is-
ics ecommended by he Na ional Ins i u e o S and-
a ds and Technology [9, p. 2]. These cha ac e is ics a e
he pilla s o wha is called cloud compu ing. The es-
igna ion o one elemen will a ec he whole cloud
compu ing model. The i s ou cha ac e is ics a e
esponsible o he lexibili y o he solu ion in e ms
o he a ailabili y o IT esou ces. Measu ed se ice
allows cha ging cus ome s acco ding o usage.
2.3 Se ice Models
So wa e as a se ice (SaaS) is de ined as he capabili y
p o ided o he consume o use he p o ide ’s applica-
ions. Recipien decides which package o p og ams
o e ed by he supplie will se e his needs. The SaaS
model allows en ep eneu s o pay only o he so wa e
ea u es hey ac ually un and cha ge o ac ual ime
o use. The so wa e package dedica ed o he indi idu-
al cus ome ope a es, main aining an adequa e le el
o da a secu i y, in addi ion o o he so wa e packages
on a sha ed se e . The clien shall de e mine he num-
be o employees who ha e access o he sha ed p o-
g ams. The numbe o employees and he ac ual usage
a e he basis o calcula ing he subsc ip ion ee. So -
wa e as a Se ice emo es he need o handle he ins al-
la ion, se -up and o en daily up keep; educes cos
o implemen a ion o new so wa e; gi es an ad an age
o speed implemen a ion and access o he la es so -
wa e. Because he so wa e is hos ed emo ely, use s do
no need o in es in addi ional ha dwa e and pu chas-
ing so wa e licenses.
Pla o m as a se ice (PaaS) is de ined as a compu ing
pla o m being deli e ed as a se ice o he cus ome
deli e ing ia he i ual wo k en i onmen , among
o he s, ope a ing sys ems such as Windows Azu e c e-
a ed by Mic oso o Google's Ch ome OS. Cus ome s
can use he in as uc u e and de elopmen ools hos ed
by a se ice p o ide o build hei own applica ions
and wo king en i onmen . This model deploys on o he
cloud in as uc u e consume -c ea ed o acqui ed ap-
plica ions wi hou he need o pu chase ha dwa e
o so wa e ins alla ion. PaaS se ice is a ailable o any
use ega dless o he used end s a ion. In addi ion,
he clien is no in ol ed in he main enance o he
ope a ing sys em i sel , upda ing, backup, e c.
The cus ome is o only main ain and ake ca e o sel -
c ea ed o ins alled applica ions. The ad an ages
o using PaaS se ices include: no need o pu chase
a comme cial license o ope a ing sys ems, no need
o upda e ope a ing en i onmen s by he se ice p o-
ide , access o he la es e sions o ope a ing sys ems
a no addi ional cha ge and he possibili y o using he
wo k en i onmen on di e en compu e s.
In as uc u e as a se ice (IaaS) is o p o ide
he ha dwa e unde s ood as RAM, disk space, compu -
e pe o mance, e c. o e he In e ne in he o m
o i ual esou ces scaled acco ding o he needs o he
ecipien . Resou ces ha a e he supplie p ope y will
be i ually sepa a ed wi hin he physical de ices unde
his con ol. The owne o he ope a ing en i onmen ,
ins alled p og ams and p ocessed da a in his se ice
model, is he ecipien . This implies a need o main e-
nance and upda ing o so wa e. The ad an ages
o using IaaS se ices emo es he ha dwa e pu chase
cos s and main enance cos s o ne wo k de ices, p o-
ides he abili y o quickly expand i ual ha dwa e
esou ces (i necessa y) and he abili y o educe i ual
ha dwa e esou ces, and he sa ing o o ice space
de e mined by he lack o need o a complex se e .
Table 1 p esen s a summa y o he desc ibed cloud
compu ing se ices and o ganizes issues o owne ship
and esponsibili y o he indi idual elemen s as he
subjec s o se ices. Sign (-) indica es ha he con ol
o he esou ce has he cus ome , sign (+) indica es ha
he esou ce is unde p o ide con ol, while he sign
(+/-) indica es ha he esou ce is unde common con-
ol o bo h en i ies.
Table 1 p esen s he dis ibu ion o esponsibili y o IT
in as uc u e be ween he p o ide and he ecipien .
In each o he lis ed cases, he company as a cus ome
bea s he esponsibili y o he da a. Responsibili y o
in as uc u e di e s, depending on he a ian . Cus-
ome mus eckon wi h he ac ha he g ea e
he esponsibili y is shi ed o he p o ide , he g ea e
will be he cos o he se ice.
Cloud Compu ing as a Tool o Imp o ing Business Compe i i eness 79
2.4 Deploymen Models
Public Cloud is he mos popula and leas engaging
cus ome ypes o cloud compu ing. The cloud in a-
s uc u e is p o isioned o open use by he gene al
public. The public use s may pay ixed mon hly ees
o se ices hey u ilize. Examples o companies o e -
ing his ype o se ice include Google, Mic oso ,
Amazon and leading Polish g oups: One .pl G oup
and ATM SIS.A. The e a e also ee e sions o public
clouds such as al eady men ioned mailboxes. Google
has gone one-s ep u he and made use s o Gmail
mailboxes ools o c ea e a ne wo k o o ice docu-
men s, Web pages and i ual disks. Ano he ea u e
o he public cloud, whose p esence should be awa e
o , is he ac ha se ices can be p o ided by mo e
han one en i y, e.g. one company p o ides i ual
ha dwa e, ano he ope a ing pla o m, and e en o he
applica ions used by he cus ome .
Taking e e y hing in o accoun , Public Cloud is a o m
o cloud compu ing, which is he mos con enien
o he cus ome . Recipien p ocu es ce ain se ices,
pays a ee and h ough In e ne connec ion which is
equi ed in his case, s a s using he se ice. Public
Cloud combined wi h IaaS, PaaS and SaaS package
se ices is mos bene icial o he company. Using his
combina ion ees he business om ha ing o pu -
chase, se ice and main ain ha dwa e and o pay
o so wa e licenses. I sa es o ice space, elec ici y
and akes ad an age o many o he bene i s ha will be
p esen ed la e in he a icle.
Why do no all bene i om his kind o echnology?
The e a e many exis ing easons in he li e a u e bu
om he business poin o iew he mos impo an
seems o be a conce n o sa e y da a p ocessing, espe-
cially classi ied in o ma ion. The e o e, secu i y con-
side a ion may be subs an ially esponsible o he
concep o p i a e, common and hyb id clouds.
P i a e cloud is a solu ion o he p oblems o da a secu-
i y in cases whe e he company does no wan o can-
no p ocess hem ou side ully con olled by hem da a
cen e s. Mo e speci ically, P i a e Cloud sol es legal
issues. Physical secu i y o da a and p o ec ion agains
hacke a acks a e a sepa a e issue.
P i a e cloud in as uc u e is ope a ed solely o
a single o ganiza ion whe he managed in e nally o by
a hi d-pa y and hos ed in e nally o ex e nally. F om
a echnical poin o iew, he pu chase and cons uc ion
o ha dwa e in as uc u e p i a e cloud is no much
di e en om he classic se e . The di e ence lies
in he laye o so wa e ha uns on he a ailable ha d-
wa e.
I is easy o no ice ha his ype o solu ion is no a ail-
able o any o ganiza ion. Unde aking a p i a e cloud
p ojec equi es a signi ican le el and deg ee o en-
gagemen and equi es he o ganiza ion o ha e su i-
cien unds and skilled employees, which is una ailable
o many companies, such as in he SME. The e o e,
hese solu ions a e ypically used by la ge companies
whe e he scale o p ojec s linked o he legal p o ec-
ion o da a p ocessed is economically jus i ied.
Howe e , he ques ion is i his ype o p i a e en e -
p ise has a chance o ac ually p o ide a highe le el
o secu i y agains hacke s o na u al disas e s ha can
lead o i e e sible da a loss. Fo la ge o ganiza ions,
such as mul ina ional co po a ions, whe e he e a e
adequa e inancial esou ces, pe haps i does. In o he
cases, companies a e no able o p o ide be e secu i y
han ex e nal supplie s.
Cloud compu ing p o ide s in es in a gian cloud da a
cen e s, o en cos ing $ 500 million o mo e. In 2008,
Google spen on he cons uc ion o da a cen e s o
$ 2.3 billion. Economies o scale a e he key ac o
applied by cloud p o ide s. Huge o de s o equipmen ,
ene gy, as well as s a egic loca ion o da a cen e s
Table 1. Dis ibu ion o con ol in models o cloud compu ing se ices
(sou ce: [6, p. 6])
Speci ica ion Classic model IaaS PaaS SaaS
Da a – - - -
Applica ions - - - +
Ope a ing en i onmen - - + +
Vi ual machine - +/- + +
Se e s - + + +
Da a wa ehouses - + + +
Ne wo k +/- + + +
80 Michał Wiśniewski
a ound he powe plan , da a cen e ope a o s can coun
on big discoun s o 60% and mo e. An example is Am-
azon, which in 2008 paid abou 90 million dolla s
o 50 housand SGIRackable se e s. In no mal ad-
ing, hey cos 215 million dolla s. [12]
We can likely ind ou ha a single o ganiza ion is
no able o gua an ee a highe le el o secu i y o hei
da a due o he unds in es ed by cloud se ice p o id-
e s, economies o scale de e mined by he amoun
o cus ome s and duplica ion o da a cen e s in di e en
pa s o he coun y o he wo ld. In addi ion, he e a e
issues ela ed o he expe ience o employees in sa e y
and inno a ion and mode ni y solu ions.
In conclusion, p i a e cloud is echnically and cos ly
neighbo ing he classical da a p ocessing. The whole
p ocess and he esponsibili ies and cos s emain a he
clien side. Howe e , p i a e cloud gains mo e e icien
use o ne wo k esou ces (se e s, disk a ays, e c.),
and e ains con ol o classi ied da a wha is impo an
o he cu en legisla ion easons.
Communi y cloud is de ined by NIST as in as uc u e
p o isioned o exclusi e use by a speci ic communi y
o consume s om o ganiza ions ha ha e sha ed con-
ce ns (e.g. mission, secu i y equi emen s, policy
and compliance conside a ions). I may be owned,
managed and ope a ed by one o mo e o he o ganiza-
ions in he communi y, a hi d pa y, o some combina-
ion o hem, and i may exis on o o p emises [10,
p. 3]. Thus, communi y cloud is in a sense a kind
o p i a e cloud connec ing en i ies wi h a simila p o-
ile ha ag ee o sha e echnical esou ces and, in some
cases, he da a. O ganiza ions ha canno a o d
o build a p i a e cloud a e in es ing in communi y
cloud, and because he law does no allow o he p o-
cessing o da a cen e s ha a e no unde con ol o he
company. As he numbe o pa icipan s in such a p o-
jec is much smalle han in he case o public cloud,
pa o he economic bene i s esul ing om economies
o scale is no possible o achie e. Howe e , compa ed
wi h he p i a e cloud cons uc ion and main enance
he cos s a e much lowe .
Hyb id cloud is a composi ion o wo o mo e dis inc
cloud in as uc u es (p i a e o public) whe e some
o he da a a e p ocessed in he public cloud and he
classi ied da a in he p i a e cloud. Hyb id cloud seeks
o deli e he ad an ages o limi ing he capaci y
o p i a e clouds, which allows ope a ing in acco dance
wi h he law and po en ial cos sa ing which is smalle
han he cos s o managemen da a o public clouds.
Despi e he deployed a ian o cloud compu ing se -
ices, he main p oblem a ises om he need o ech-
nological s anda ds. This leads o a si ua ion whe e he
clien has limi ed abili y o change se ice p o ide .
This ca ies a isk o dependence on single supplie
companies, who may impose in la ed p ice o he se -
ice.
2.5 Selec ion o cloud compu ing se ices
The main mo i e in a o o he deploymen o cloud
compu ing solu ions in o ganiza ions bo h p i a e
and public is economic conside a ions. En e p ises
p ocessing he da a in a classical way mus hemsel es
ake ca e o all he ela ed aspec s o he p ocess such
as: secu i y, mode niza ion o IT in as uc u e,
and so wa e licenses. Cloud compu ing p o ides many
oppo uni ies o cos sa ings o companies and indi-
idual use s. I should be no ed, howe e , ha in i s
cu en o m, he cloud compu ing is no a pe ec solu-
ion, and as each new solu ion has i s d awbacks,
i elimina es pa o he known isks, bu also c ea es
new ones.
The au ho s o he a ious s udies ag ee ha he main
ad an ages o cloud compu ing include sa ing, pe o -
mance, scalabili y, a ailabili y, eliabili y, ime o im-
plemen a ion o new unc ionali y. Launching new
Google Apps se ices a he Uni e si y o Wa saw is
an example o a much as e launch o IT se ices.
Wi hin h ee mon hs o he se ice unning he UW
s uden s we e p o ided wi h he access o email, s o -
age o documen s on he web, c ea ion web pages,
blogs and he use o o ice so wa e package oge he
wi h he ans e o exis ing s uden mailboxes om
he uni e si y se e [14]. The launch o a new mul i-
media se ice newspape Rzeczpospoli a may se e as
a second example. The si e was buil en i ely in he
cloud wi hin h ee weeks. Fo compa ison, he con-
s uc ion ime o he same se ice in he ba o m
would ake om 6 weeks o 4 mon hs [6, pp. 20–22].
Cloud compu ing secu i y was no lis ed delibe a ely
among he ad an ages because he opinions di e on
his ma e . Some au ho s conside cloud solu ions a
sa e han adi ional p o ec ion mechanisms wha
in iew o applied ele an p ocedu es and echnical
solu ions should be conside ed as mo e e ec i e.
On he o he hand, he p i acy and legisla ion canno be
unde es ima ed.
Cloud Compu ing as a Tool o Imp o ing Business Compe i i eness 81
Figu e 2. Elemen s o he IT in as uc u e o he analyzed online s o e
(sou ce: own elabo a ion)
Use s s o ing and p ocessing da a in he cloud do no
always know whe e hei da a a e physically loca ed
and who and on wha basis has access o hem. Da a
b each is a big conce n in cloud compu ing. Cus ome s
wo y i p o ide cloud compu ing se ices in o ms
hem abou any b each o secu i y in hei da a cen e s.
In addi ion, some companies o e ing cloud compu ing
se ices has egis e ed hei o ices ou side he Eu ope-
an Union and a e no subjec ed o EU law, which may
lead o lowe s anda ds o da a secu i y.
Disad an ages associa ed wi h unce ain ies abou he
sa e y o he cloud compu ing se ices a e mul iplied,
bu limi a ions as dependence on so wa e p o ide s,
he lack o eadiness o employees o use he cloud, IT
specialis s’ eluc ance o po en ially dange ous changes
o hei employmen , in e ne links wi h insu icien
bandwid h.
Conside ing he ad an ages and disad an ages, oppo -
uni ies and h ea s ela ed o he p ocessing o da a
in he cloud should compu e he ype o cloud (public,
p i a e, sha ed, hyb id) and he ange o se ices o -
e ed should be aken in o accoun . Mos o hese bene-
i s a e a ailable i he company chooses IaaS, PaaS,
SaaS models. The adop ion o hese models is mos
p o i able o he company p o iding mo e lexibili y,
lowe cos s and ease o use. In con as , i is no so
e iden in he case o p i a e cloud. In his case,
he solu ion, in e ms o cos s and esponsibili ies, di -
e s sligh ly om he classical da a p ocessing solu-
ions. Because he da a cen e is unde he con ol
o he company i dissol es he p oblem associa ed wi h
legal p o isions.
How o de e mine whe he ansi ion om he classical
model o cloud compu ing model is cos -e ec i e
and which se o cloud se ices, de e mined by a com-
bina ion o a ailable se ice models, deploymen mod-
els and ypes o clouds, is he bes o he o ganiza ion?
I seems ha in o de o answe his ques ion, i is nec-
essa y o de elop a me hod o balancing he cos s
and expec ed bene i s o bo h classical and cloud com-
pu ing solu ions. Such a me hod should ake in o ac-
coun no only he cos o cons uc ion and equipmen
o he se e bu also he speed o deploymen o new
unc ionali ies, which ha e been shown on he example
o a Rzeczpospoli a newspape mul imedia se ice,
he cos o quali ied s a employing, ope a ing cos s
and he cos s o disposing o was e o unnecessa y IT
equipmen .
3 Case S udy
This case s udy demons a es how company is planning
wi hin exis ing ope a ions he c ea ion o an online
s o e selling spo ing goods. The online s o e ac i i ies
we e o be main ained o h ee consecu i e yea s. A -
e his ime, he decision o con inue o discon inue
ope a ions migh be made. The company has al eady
c ea ed he necessa y IT in as uc u e including i s
own da a cen e and quali ied s a . To s a a new p o-
jec , a company mus pu chase addi ional ha dwa e. All
new ha dwa e componen s a e shown in Figu e 2. To
ensu e con inui y o se ice a ailabili y and o educe
ou age du a ion each o he elemen s will be duplica ed
as p esen ed in Figu e 2. P ojec managemen assumes
a s able IT in as uc u e load wi hou load su ges.
82 Michał Wiśniewski
Table 2. Cos o pu chasing new IT in as uc u e elemen s needed o launching an online s o e
(sou ce: own elabo a ion)
Speci ica ion Uni p ice Numbe
o i ems In e nal solu ion
Fi ewall 4000 2 8,000
Equi alen load 14000 2 28,000
Ne wo k se e 8500 2 17,000
Applica ion se e 8500 2 17,000
Da a se e 8500 2 17,000
To al - - 87,000
F ee so wa e will be used o minimize he building
cos s o an online s o e. Necessi y o pu chase a new
ha dwa e and he unce ain y whe he he shop will
con inue o exis a e he assumed ime begs he ques-
ion. The ques ion may ye be aised in ano he case
whe he he cos o new IT equipmen and launching
o he p ojec in he company’s da a cen e will be p o -
i able. Is he lease o i ual s o es an ad an ageous
solu ion in he cloud compu ing o mula?
P ojec managemen deciding o launch an online s o e
in hei own da a cen e mus bea he cos s p esen ed
in Table 2.
The cos o pu chasing he necessa y elemen s o he IT
in as uc u e o launching an online s o e in hei own
da a cen e is a he le el o abou 87,000 PLN. Assum-
ing a h ee-yea pe iod o he p ojec , he cos is
2416,67 PLN pe mon h. The o al amoun mus be
paid in ad ance. The calcula ion applies only o he
pu chase o necessa y new in as uc u e elemen s.
Cos s no aken in o accoun include:
powe supply o pu chased equipmen ,
cooling o pu chased equipmen ,
se e ack space,
in e ne connec ion ( o example, an addi ional link
10 Mb–3000PLN pe mon h),
s a emune a ion (adminis a o s and echnicians),
s a aining.
These cos ca ego ies a e no included in u he analy-
sis because hey a e al eady paid by he company,
which main ains i s own da a cen e s. The inc ease
in ope a ing cos s due o he pu chase o new IT
equipmen makes only small pa o o al cos s
and would be conside ed insigni ican . The e o e, we
a e no going o analyze hose cos s any mo e.
Summing up he cos o launching, an online s o e
in he en e p ise da a cen e , he adop ed assump ions
ha mon h abou 2,400 PLN a a h ee-yea e m adds
up o abou 87,000 PLN. Howe e , he e is a possibili y
ha a e his planned pe iod he p ojec will be liqui-
da ed as unp o i able. I his happens, he company
emains wi h unused ne wo k equipmen ha should be
main ained o o bea he cos s o disposal.
An al e na i e solu ion o elimina e he p oblem
o unnecessa y IT equipmen is o launch an online
s o e in he cloud compu ing en i onmen . S a ing
a p ojec wi hin cloud compu ing o mula o e s cos
sa ings because o “pay-as-you-go” se ices wi h
no con ac s, and o en no sign-up ees and jus we pay
o wha we use, e.g. RAM, CPU, ha d d i es.
A company o un an online s o e in cloud compu ing
ins ead o buying physical ha dwa e pu chases i s i -
ual coun e pa s. Table 3 p esen s a summa y o i ual
ins ances o e ed by Amazon, Mic oso and Polish
company. Unlike Amazon and Mic oso , Polish com-
pany does no o e p ede ined ins ance bu allows
con igu a ion o a i ual se e acco ding o clien ’s
needs. In case o Amazon Medium Ins ance, i co e-
sponds o he physical se e needed o he planned
c ea ion o an online s o e. And Mic oso La ge In-
s ance mee s he equi emen s o he p ojec . To ensu e
he compa abili y o i ual ins ances o Polish com-
pany, iden ical pa ame e s o i ual se e s as o he
selec ed ins ance Amazon shall be adop ed. Ins ances
eplace i ual IT in as uc u e elemen s such as
a ne wo k se e , applica ion se e and da a se e .
I is necessa y o ake ou six i ual ins ances o e-
place six physical se e s o assumed IT s uc u e.
To build a comp ehensi e IT in as uc u e, a se ice
mus be pu chased o eplace he i ewall and se e
load equi alen . Table 4 gi es he mon hly en al cos
o such se ices in compa able companies assuming
ha he se ices a e s ill a ailable o all en ed in-
s ances.