TELKOM
NIKA
, Vol.11, No
.3, March 2
0
1
3
, pp. 1646 ~ 1651
ISSN: 2302-4
046
1646
Re
cei
v
ed
No
vem
ber 3, 20
12; Re
vised Janua
ry 2
6
, 20
13; Accepted
February 7, 2
013
A SaaS Development Platform Based on Cloud
Computing
Qingpeng Z
e
ng*
1,a
, Shuixiu Wu
2,b
1
School of Infor
m
ation En
gi
ne
erin
g,
NanC
ha
ng Un
iversit
y
, NanC
ha
ng, Ch
ina
2
School of Co
mputer an
d Informatio
n
Engi
n
eeri
ng, Jian
g
x
i
Normal U
n
iv
ersit
y
, NanC
ha
n
g
, Chin
a
*Corres
p
o
ndi
n
g
author, e-ma
i
l
: zengq
in
gpe
n
g
@nc
u
.edu.c
n
a
, w
u
sh
ui
xi
u@
sina.com
b
A
b
st
r
a
ct
Using Jav
a
techno
logy to res
earch an
d dev
elo
p
a SaaS d
e
vel
o
p
m
e
n
t and mainte
na
nc
e platfor
m
w
h
ich is b
a
sed
on cl
oud c
o
mputin
g, it incl
ud
es many
modu
les suc
h
as the
clou
d co
mp
uti
ng l
oad
netw
o
r
k
,
application server, devel
opm
e
nt framework, cl
oud com
p
uting dev
elopm
e
nt engin
e, oper
ation
and
ma
inte
nanc
e s
ystem a
nd so
on. the p
l
atfor
m
ca
n prov
id
e
a co
mpl
e
te n
e
tw
ork client softw
are deve
l
op
men
t
sche
m
e for th
e dev
elo
p
m
e
n
t
of softw
are
and
integr
atio
n, it can red
u
c
e the us
er
’
s
requ
est to Sa
aS
techno
lo
gy, an
d it ca
n h
e
lp t
he us
er to
achi
eve o
per
ation
service
of pro
d
u
ct on th
e Inte
rnet. at the s
a
me
time, the p
l
atfo
rm as a d
e
vel
o
pment tool ca
n
be used as
a pow
erful sup
p
l
e
ment of core
electro
n
ic dev
i
c
es,
hig
h
-en
d
ge
ner
al chi
p
an
d bas
ic softw
are products.
Ke
y
w
ords
: Cl
oud C
o
mputi
n
g
,
SaaS, Middle
w
are
Copy
right
©
2013 Un
ive
r
sita
s Ah
mad
Dah
l
an
. All rig
h
t
s r
ese
rved
.
1. Introduc
tion
Clou
d com
p
u
t
ing is the a
b
ility to acce
ss
a p
ool of
comp
uting reso
urce
s owned an
d
maintaine
d
b
y
a third p
a
rty
via the Intern
et. It isn’t a n
e
w te
chn
o
log
y
but a ne
w
way of delive
r
ing
comp
uting re
sou
r
ces ba
sed
on
lo
ng existing
te
ch
nol
ogi
es su
ch
as se
rver virtualizatio
n.
the
“clo
ud
” is
co
mposed
of h
a
rd
wa
re, sto
r
age, net
wo
rk
s, interfa
c
e
s
,
and
se
rvice
s
that provid
e
the
mean
s throu
gh whi
c
h users a
c
ce
ss the infra
s
tr
u
c
ture, com
puti
ng power, a
pplication
s
, and
servi
c
e
s
on
d
e
mand
and i
n
depe
ndent of
locatio
n
. Clou
d com
puting
usu
a
lly involves the tran
sfe
r
,
stora
ge, and
processin
g
of in
formatio
n on the pro
v
ider’s inf
r
a
s
tructu
re, which is outsi
de
the
cu
stome
r
’s control
[1-2].
Clou
d co
mp
uting ha
s fo
ur big
ch
ara
c
teri
st
ics whi
c
h a
r
e ta
kin
g
the network a
s
the
cente
r
, to provide service way, high extende
d hi
gh reliability, reso
urce pooling ,transpa
rent and
so on, with t
he app
rea
r
in
g of cloud
co
mputing,
it bring the IT re
sou
r
ces
with
the operatio
nal
con
d
ition
s
. Cloud co
mputin
g includ
es three kind
s
of mode whi
c
h a
r
e the I
aaS (Infrastructu
r
e a
s
a se
rvice, Infrastructu
re t
hat is Se
rvi
c
e),
Pa
aS (Platform a
s
a Service Pl
atform, nam
ely
servi
c
e
s
), Sa
aS (Software
as a Service
,
Softwar
e that is Service), and the core technol
ogy
o
f
the mode
s involves ha
s huge difference. The
three level
s
of cloud
co
mputing hav
e not
inevitable rel
a
tion on t
e
ch
nically, but in
the eye
s
of t
he technol
og
y develop tre
nd an
d p
r
a
c
tical,
the relationship of the three levels
will be more and more cl
osely.
The e
a
ch pa
rt of the Clo
u
d
com
puting
i
n
clud
es many prog
ram
m
ing
langu
age,
o
p
e
rating
system, data
base, Web
server, ag
ree
m
ent and ap
plicatio
n pro
g
r
ammin
g
inte
rface
(API). The
key is to identify which cl
o
ud se
rvice re
ally su
itable for their intern
al system
s, application
s
and
skill
s. Amazo
n's el
asti
c ca
lculatio
n clo
u
d
, Google a
p
p
licatio
n engi
ne and
Wind
ows Azu
r
e a
r
e
three typical
repre
s
e
n
tatives.
Amazo
n's EC2 provide
s
cu
stome
r
s with a va
riety of software choi
ce, such as Windo
ws
Server, Op
en
Solari
s and
seven Linux v
e
rsi
on,
MySQL, SQL Server, and som
e
develop
me
nt
environ
ment
su
ch a
s
Oracl
e
11 g datab
a
s
e, Ja
va, JB
o
ss, Ruby on
Rail
s and
so
on [3, 7, 12].
The
spe
c
ialty of Googl
e is
easy to u
s
e f
o
r u
s
. The
ap
plicatio
n engi
ne can h
e
lp
use
r
s t
o
use G
oogl
e's indepe
nde
nt to develop databa
se a
n
d
other infra
s
tru
c
ture soft
ware; and it can
throug
h the A
P
I to use
ca
ching, mirro
r
,
mail an
d othe
r services. Py
thon is the o
n
l
y prog
rammi
ng
langu
age
whi
c
h is
su
ppo
rted, but Goo
g
l
e
is al
so goi
n
g
to sup
port
other p
r
og
ra
mming lan
g
u
age
s
in the future.
Evaluation Warning : The document was created with Spire.PDF for Python.
TELKOM
NIKA
ISSN:
2302-4
046
A SaaS Deve
lopm
ent Platform
base
d
on
Cloud
Com
puting (Qin
gpe
ng Zeng
)
1647
The Wind
ows
Azure
an
d
Azure servi
c
e
pl
atform has a
relatio
n
with
th
e
M
i
cro
s
oft'
s
internal
deplo
y
ment of ente
r
pri
s
e
softwa
r
e se
rie
s
. The
Azure incl
ude
s SQL Se
rver of trustee
s
hi
p
versio
nst, Dy
namics CRM
and net serv
ice, it is
developed by usin
g Visual Studio and the.ne
t
frame
w
ork. The Microsoft
says, Azu
r
e will s
upp
ort the open protocol (HTTP, REST, SOAP and
XML) and the
Micro
s
oft pro
g
rammi
ng lan
guag
e (Ecli
p
se, Ruby, PHP and Python).
2. The Main Work
Clou
d co
mp
uting build
s heavily on
capa
bilities available t
h
rou
gh
seve
ral core
techn
o
logie
s
and capa
bilities [4-6] :
(1) Web app
lication an
d servi
c
e
s
. Software a
s
a service (Sa
a
S
)
and platform as a
servi
c
e
(Paa
S) are unthi
n
k
abl
e with
out
We
b ap
p
lica
t
ion and
Web
se
rvice
s
te
chnolo
g
ies. S
a
a
S
offering
s are
typically implemente
d
a
s
Web
ap
pl
ication
s
, whil
e PaaS offering
s p
r
ovi
d
e
developm
ent and runtime
environ
ment
s for Web appl
ic
ation
s
and servi
c
e
s
. For infrast
r
u
c
ture
as
a se
rvice (Ia
a
S
) offering
s, admini
s
trato
r
s typica
lly im
plement a
s
so
ciated
servi
c
e
s
and APIs,such
as the man
a
g
e
ment acce
ss for custom
e
r
s,
u
s
ing Web appli
c
ation/
se
rvice
te
chnol
ogie
s
.
(2) Virtu
a
lization IaaS offering
s. The
s
e
tech
n
o
logie
s
have virtualization tech
niq
ues at
their very he
art; becau
se
PaaS and Sa
aS servi
c
e
s
a
r
e u
s
ually bui
lt on top of a sup
portin
g
IaaS
infrast
r
u
c
ture,
the importance of virtualization
also extends to these servi
c
e
models. In
the
future, we e
x
pect virtuali
z
ation to de
velop fr
om
virtualize
d
servers toward comp
utatio
nal
resou
r
ces tha
t
can be u
s
ed
more readily for executin
g SaaS servi
c
e
s
.
(3) O
n
-dema
nd self
-servi
ce. Use
r
s
ca
n ord
e
r an
d
manag
e se
rvices
without
human
intera
ction with the service provid
er,
using,
for example, a Web po
rtal and mana
ge
ment
interface.Pro
v
isionin
g
an
d de-provi
s
i
oning of
se
rv
ice
s
an
d as
so
ciat
ed
r
e
so
ur
ce
s o
c
cur
automatically at the provide
r
.
(4) Ubiquito
u
s
netwo
rk acce
ss. Clo
ud
servi
c
e
s
are
acce
ssed via the network (usually
the Internet),
usin
g stan
dar
d mech
ani
sm
s and p
r
oto
c
o
l
s.
(5)
Re
sou
r
ce
pooling. Co
mputing reso
urces
used to provid
e the clou
d serv
ice a
r
e
reali
z
ed u
s
in
g a homog
en
eou
s infra
s
tru
c
ture th
at’s share
d
betwee
n
all servi
c
e u
s
ers.
(6) Ra
pid
el
asticity.
Re
source
s can be
sc
al
ed u
p
and
do
wn
rapid
an
d
elasti
cally.
resou
r
ce/se
r
vice usage is
con
s
tantly metered,
supp
orting optimi
z
ation of reso
u
r
ce u
s
a
ge,u
s
age
repo
rting to the cu
stome
r
, and pay-as-y
ou-g
o
bu
sine
ss m
odel
s.
The main ta
sk of our pl
atform is u
s
in
g the Java technolo
g
y to resea
r
ch and d
e
velop a
platform
whi
c
h is b
a
sed on
the SaaS’s
d
e
velopm
e
n
t, operation a
n
d
maintena
nce
,
it include
s fi
ve
module
s
whi
c
h
a
r
e clou
d comp
uting
lo
ad
net
work,
clou
d comput
ing en
gine, a
pplication
server
and develop
ment framework, the
se
rvice delivery pl
atform, opera
t
ion and main
tenan
ce syst
em.
The overall lo
gical
stru
cture as sho
w
n in
figure 1:
Figure 1. Fra
m
ewo
r
k of
SaaS Develo
p
m
ent Platform base
d
on
Clou
d Com
p
u
t
ing
Evaluation Warning : The document was created with Spire.PDF for Python.
ISSN: 2302-4
046
TELKOM
NIKA
Vol. 11, No
. 3, March 20
13 : 1647 – 1
652
1648
2.1. Cloud Computing Lo
ads Ne
t
w
o
r
k
Clou
d comp
u
t
ing loads net
work is form
e
d
ma
inly by the clou
d com
puting re
sou
r
ce pool,
infrast
r
u
c
ture
manag
eme
n
t
serve
r
an
d
appli
c
ation
software.
Clou
d co
mputing
resou
r
ce po
o
l
is
formed
by a grou
p of CP
U which
ca
n
reali
z
e aut
o
m
atic supply
and recovery of beforeh
and
cu
stom-built and mem
o
ry
resour
ce
s, it con
c
rete include
s se
rver
node e
quipm
ent and
singl
e
stora
ge d
e
vices. Be
cau
s
e
of many characteri
stic
s of
the clo
ud pl
atform, it need
s a big
re
sou
r
ce
pool to reflect, such as dy
namic mig
r
ati
on, load
bala
n
ce an
d high
availability,
so we u
s
e two
serve
r
s and a
disk a
r
ray to con
s
titute.
The infra
s
tru
c
ture
mana
g
e
ment serve
r
of
the Cloud
comp
uting i
s
a de
dicate
d se
rver,
whi
c
h b
a
si
c function i
s
to installati
on,
ope
ratio
n
and
man
ageme
n
t of the entire
clou
d
infras
t
r
uc
ture toolk
i
t.
The Clo
ud computing a
p
p
licatio
n soft
ware
(b
asi
c
packa
ge) i
s
based on
We
b2.0,and it
doe
s the re
sou
r
ces
sup
p
ly and recovery (su
c
h
as Web se
rver, Web a
pplication se
rver,
databa
se server,
dire
ctory servi
c
e
s
,
et
c.
) acco
rdin
g to the servi
c
e reque
st.
2.2. Cloud Computing Application E
ngine
The sea
r
ch engine
s is a deployment an
d operat
io
n platform whi
c
h
is built on
the basi
s
of the traditio
nal op
eratin
g
system o
r
th
e appli
c
at
ion
of PaaS, it throug
h the Int
e
rnet to p
r
ovi
d
e
appli
c
ation d
e
velopme
n
t, testing, de
plo
y
ment and
h
o
sting
se
rvices for th
e de
velopers, an
d it
throug
h the
way of pl
atform servi
c
e to
delivery
a
ppl
ication i
n
fra
s
t
r
uctu
re, the
d
e
velope
rs ne
ed
not to care about the complicated tech
nical ar
chitecture an
d details of
deployment
and
impleme
n
tation, it can d
e
velop cl
oud
comp
ut
ing appli
c
ation whi
c
h
ha
s chara
c
te
risti
c
s
of
automatic e
n
j
o
y load bala
n
ce, ela
s
tic e
x
pansi
on, an
d disa
ster tol
e
ran
c
e. ap
pli
c
ation en
gine
is
mainly forme
d
by the ap
plicatio
n ho
sting env
ironm
ent, clou
d a
pplication mi
ddle
w
are, m
any
tenants d
a
ta
base, the environm
ent of applic
ation developm
ent and the
environm
en
t of
appli
c
ation te
st [8-11].
The a
ppli
c
ati
on ho
sting
e
n
vironm
ent i
s
t
he
co
re
of the ap
plication of the
engin
e
comp
one
nts,
it suppo
rts the depl
oym
ent of the a
pplication,
cu
stody and o
peratio
n, and
it
provide
s
the
monitori
ng of the app
licati
on, the sch
e
duling of the
resou
r
ce and
isolate se
cu
rely.
The a
pplication which is
d
eployed i
n
ho
sting e
n
vi
ron
m
ent can
sha
r
e the
re
sou
r
ce
s of the Pa
aS
platform, it can be divided into two kind
s of hosting environme
n
t wh
ich are the sh
ared ha
rd
wa
re
and the
sh
ared platfo
rm a
c
cordi
ng to t
he sha
r
ing l
e
vel of the re
source
s.
share
d
ho
sting
se
rver
hard
w
a
r
e
env
ironm
ent is b
u
ilt based o
n
the virtual
m
a
chin
e an
d tra
d
itional a
ppli
c
ation contain
e
r,
and ea
ch virt
ual ma
chine
only can d
epl
oy a applic
ation example,
relying on th
e
virtual machi
ne
to impleme
n
t appli
c
ation of
coa
r
se pa
rticle si
ze
seg
r
e
gation, the a
p
p
licatio
n whi
c
h is d
eployed
in
hostin
g
environment only can s
hare se
rver and ope
rating system,
and each ap
plicatio
n nee
ds
to be indep
e
ndent ap
plica
t
ion contai
ne
r, middl
e
w
are and data
b
a
s
e, the traditi
onal ap
plicati
on
without modif
i
cation ca
n direct depl
oym
ent; shar
in
g platform ho
sting environm
ent is a kind of
the fine particle size i
s
olati
on techni
que
extensio
n ap
plicatio
n se
rver, deploye
d
in its application
not only sharing platform
hard
w
a
r
e
an
d ope
rating
system, a
ppli
c
ation
of sh
a
r
ing pl
atform
of
contai
ners, middle
w
a
r
e and
data
b
a
s
e,
hosting
e
n
v
ironme
n
t through the
sa
ndbox techn
o
log
y
appli
c
ation to
reali
z
e fine g
r
anul
arity, iso
l
ation,
and th
e appli
c
ation
of use of the
resou
r
ces of t
he
real
-time mon
i
toring an
d accurate mea
s
u
r
eme
n
t.
Clou
d appli
c
ation middle
w
are se
rvice
s
prov
id
es b
u
sin
e
ss ind
e
pend
ent clo
u
d
applie
d
basi
c
fun
c
tio
n
for the
ho
sting
clou
d a
pplicati
on, a
nd compa
r
e
d
to the tradit
i
onal e
n
terp
ri
se
middle
w
a
r
e, the cl
oud
appl
ication mi
ddle
w
are ha
s m
a
ny
tenants, a
nd it ca
n pr
ovide services f
o
r
multiple deve
l
opers and m
u
ltiple applica
t
ions at t
he same time, providing the service interfa
c
e
that is based
on Web
ca
n expand fle
x
ibly acco
rdi
ng to the ap
plied load.
cl
oud ap
plicati
o
n
middle
w
a
r
e
mainly incl
ud
es
clou
d cha
nge p
r
o
c
e
s
s
servi
c
e, b
u
s
servi
c
e
s
an
d servi
c
e strate
gy,
whi
c
h is u
s
e
d
to sup
port
the application of
driven,
the application of
event driven a
nd th
e
appli
c
ation of
time period
whi
c
h are de
pl
oyed on the
application p
r
ocess of eng
ine.
2.3. Application Ser
v
er and Dev
e
lopment Frame
w
o
r
k
The ap
plication se
rver
ha
s application
star
tup, depl
oyment, online
update fun
c
ti
ons a
nd
use
r
online
state mana
g
e
ment, it mainly rea
lize t
he cent
rali
ze
d manag
eme
n
t of the server
resou
r
ces. t
he ap
plicatio
n develo
p
m
ent envir
on
ment is
use
d
for the
d
e
velopme
n
t and
deployme
nt of the application of cloud computin
g appli
c
ation o
n
the engine
, it support
s
the
developm
ent
of the spe
c
ific langu
age
of applicat
io
n of engin
e
, it provide t
he cl
ass lib
rary,
Evaluation Warning : The document was created with Spire.PDF for Python.
TELKOM
NIKA
ISSN:
2302-4
046
A SaaS Deve
lopm
ent Platform
base
d
on
Cloud
Com
puting (Qin
gpe
ng Zeng
)
1649
comp
one
nts,
appli
c
ation
s
, frame
w
ork a
n
d
debu
ggin
g
tools which a
r
e u
s
ed to de
velop the clo
ud
appli
c
ation, i
t
can
red
u
ce the thresh
old of
clou
d
appli
c
ation
develop
men
t. Developm
ent
frame
w
ork which i
s
based
on Java dev
elopme
n
t t
ools to reali
z
e S
aaS appli
c
ati
on develop
m
ent
sup
port, it in
clud
es inte
rfa
c
e devel
opm
ent, applicat
ion integ
r
atio
n, distrib
u
ted
servi
c
e a
ccess,
debu
gging,
deployme
nt, and other functions,
and
it provides with th
e operation
and
maintena
nce
system interactive ability
,
and to r
ealize to contro
l the user a
c
cordi
ng to the
function
of the software. T
he main
sol
u
tion is
a
c
co
rdi
ng to the research a
nd d
e
v
elopment of
the
Web proje
c
t modula
r
deve
l
opment and
integrate
d
fra
m
ewo
r
k Web
S
tarter, it is b
a
se
d on OSGi,
template technolo
g
y developme
n
t, su
pport the to
p
i
c, singl
e pa
ge, menu
co
nfiguratio
n, u
s
er
manag
eme
n
t can be
repl
a
c
ed, an
d mo
dule hot plu
g
, supp
ort for
multiple pag
e
technol
ogy, and
other featu
r
e
s
.
2.4. Ser
v
ice
Deliv
er
y
Plat
form
At present, the indu
stry t
o
bu
sine
ss
d
e
livery
platform is exa
c
t a
nd sta
nda
rd
definition,
different ma
n
u
facturers a
c
cording
to the
i
r un
der
stan
di
ng of in
dep
en
dent devel
op
ment of relat
e
d
prod
uct
s
. se
rvice delivery
platform is gene
rally
ref
e
rs to p
r
ovid
e uniform cl
oud appli
c
ati
on
acce
ss an
d manag
eme
n
t, suppo
rting a
pplication pro
duct
s
pre
-
sal
e
, sale and a
fter-sale se
rvi
c
e,
and it can sp
eed up the application onl
ine relea
s
e
service platfo
rm, busine
s
s delivery platform
usu
a
lly con
s
i
s
ts of appli
c
ation mana
g
e
ment, pro
d
u
c
t mana
gem
ent, use
r
ma
nagem
ent, order
manag
eme
n
t, billing and a
pplication fun
c
tion
s su
ch a
s
sup
e
rm
arke
t makes p
a
rts. This platform
busi
n
e
ss deli
v
ery of the
platform mai
n
ly r
ealize
s
the SaaS application manag
ement and
provide
s
fo
r SaaS
and cl
oud com
puti
ng
sy
stem conne
ction
abi
lity, it is the use
r
'
s
mai
n
ly use
platform. It includ
es three functio
n
s, the
first one
is th
e portal sy
ste
m
which is re
spo
n
si
ble for
the
entran
c
e
of the system
relat
ed
wo
rk,
the se
con
d
is the sy
st
em co
nfigu
r
ation which
is
respon
sibl
e for the
appli
c
a
t
ion of
the
system a
nd resource m
ana
g
e
ment, an
d th
e third
one i
s
the
use
r
se
rvice
whi
c
h is re
sp
onsi
b
le for the custo
m
er o
r
de
r related
work. The
r
e is an relation
ship
betwe
en the
m
, and it ca
n
provide it
s o
w
n
servi
c
e
fo
r othe
r SaaS
appli
c
ation a
nd the third p
a
rty
s
o
ftware.
2.5. Opera
t
ion and Mainte
nance Sy
stem
The ope
ratio
n
and maint
enan
ce sy
ste
m
can reali
z
e to manage
the user an
d buying
SaaS functio
n
, and it ha
s many functi
ons
su
ch
a
s
the time, the data qua
ntity statistics, form
orde
r, form the offer and
software o
p
e
r
a
t
ion servi
c
e.
3. Applicatio
n Analy
s
is
The software
indu
stry is the st
ate's strategic ne
w ind
u
stry,
it is the
important foundatio
n
of the national econ
omy and so
cial inf
o
rmati
z
at
ion.
promote the
software ch
ange the ori
g
inal
work i
s
the first co
ndition t
o
pro
m
ote th
e indu
stry'
s
i
nnovation
abi
lity
and gua
ra
ntee the h
eal
thy
developm
ent
of the software ind
u
st
ry. this platform
as the ele
c
tri
c
po
wer in
du
stry inform
ation
appli
c
ation system, it provides users with t
he involved in engin
eerin
g desi
g
n, project bu
dget
prep
aration, settlement m
anag
ement a
nd a seri
es
o
f
online appli
c
ation service, and it give
s full
play to the po
wer
engi
nee
ri
ng inform
atio
n data cent
er online a
pplication se
rvice busi
n
e
ss valu
e,
and it enh
a
n
ce
s the u
s
e
r
for the o
r
igi
nal soft
ware
depe
nd
s on t
h
is platfo
rm, as the el
ect
r
i
c
power indu
stry engineeri
n
g
technolo
g
y and eco
nomi
c
comp
re
hen
si
ve applicatio
n platform, so it
will solve the
electri
c
power indu
stry application sof
t
ware effe
ctiv
ely, at
the same time, it will
provide effective help for reduci
ng the enterp
r
is
e co
st and improving the utilization rat
e
of
r
e
sour
ces
.
Usi
ng the clo
ud comp
uting
platform, we can con
c
ent
rate the IT resou
r
ces to the data
resou
r
ces p
o
o
l, and the
develop
ers n
eed to ap
pl
y for acco
rdin
g to the act
ual use of IT
resou
r
ces, so
the manag
ers ca
n monito
r to the
use
of IT reso
urces,
incl
udin
g
CPU, mem
o
ry,
and mem
o
ry usa
ge. at the time, cloud computing pl
atform IT resou
r
ce
s sch
eduli
ng mana
gem
ent
and re
so
urce
recove
ry function, ca
n en
han
ce the
re
sou
r
ce usage
efficiency furthest, so it can
avoid the wa
ste of IT resource
s. acco
rding to the survey, t
he per capita PC IT reso
urces
utilization rate is less tha
n
20%, but with the
cloud
comp
uting pl
atform, it can incre
a
se the IT
resource utilization rate to 70%.
Usi
ng the clou
d com
p
u
t
ing platform
to
deploym
ent softwa
r
e
system , completely
throug
h the system
atizati
on and sta
n
dardi
zatio
n
process, whi
c
h can arran
g
e
all the work in
Evaluation Warning : The document was created with Spire.PDF for Python.
ISSN: 2302-4
046
TELKOM
NIKA
Vol. 11, No
. 3, March 20
13 : 1647 – 1
652
1650
proje
c
t group
at the same time, and it red
u
ce t
he d
eplo
y
ment of softwar
e develo
p
m
ent life cycl
e,
and it saves
the co
st of personn
el. usi
ng the cl
ou
d comp
uting pl
atform, the service life of the
averag
e serv
er is 5 to 8 y
ears, and it i
s
more
sta
b
le
, more proce
ssi
ng po
we
r. as we kno
w
, the
PC efficien
cy
is low, the f
a
ct is a
bout
1000 P
C
s
eq
ual to 10 to
15 table IBM
X3850
serv
er,
according to the 6 years de
pre
c
iation
,
it
sav
e
s 40%
I
T
cost
di
re
ct
ly
.
4. Conclusio
n
Usi
ng Java techn
o
logy to rese
arch an
d
develop a Sa
aS developm
ent and main
tenan
ce
platform which is ba
sed on
cloud
comp
u
t
ing, it
main inclu
d
e
s
four
module
s
which are the cl
o
ud
comp
uting lo
ad netwo
rk, the applica
t
ion serv
e
r
, the develop
ment frame
w
ork, the cl
oud
comp
uting d
e
v
elopment e
n
g
ine, ope
rati
on and m
a
in
t
enan
ce
syste
m
. the cloud
comp
uting lo
ad
netwo
rk i
s
be
long
s to the infrast
r
u
c
ture;
applicat
io
n server ha
s ap
plicatio
n start
up, deployme
nt,
online upd
ate function
s and use
r
onlin
e state mana
gement , it
mainly realize the centrali
ze
d
manag
eme
n
t of the server resou
r
ce
s; devel
opm
ent framewo
r
k which is used in Java
developm
ent
tools to realize SaaS appli
c
ation
d
e
velopme
n
t sup
port, it inclu
d
e
s
interface
developm
ent, application i
n
tegratio
n, di
stribute
d
service a
c
cess, debu
gging, d
eployment, a
n
d
other fun
c
tion
s, it provide
s
with the op
erati
on and m
a
i
n
tenan
ce
system intera
ctive ability, and it
can re
alize to control the
user acco
rdi
ng to
the use’s the software functio
n
;
clou
d comp
uting
engin
e
is mainly for SaaS applicatio
n whic
h ca
n
provide clo
ud com
putin
g servi
c
e
s
, and
according to
the cal
c
ul
ation amo
unt a
u
tomati
c ta
sk or
sto
r
ag
e
requ
est di
st
ributed to
cl
oud
comp
uting
server g
r
o
up
focu
s for Sa
aS softw
a
r
e provide com
puting
an
d
storage ca
pa
city;
operation an
d maintenan
ce system to realize
the user and buyin
g SaaS function manag
em
ent,
the data
qua
ntity statistics and
software
ope
ration
se
rv
ice
of relate
d functio
n
s. t
he platfo
rm
can
provide
a co
mplete net
work
clie
nt so
ftware d
e
vel
opment
sche
me for the
developm
ent
o
f
softwa
r
e a
n
d
integratio
n, it can redu
ce
the user
’
s
re
que
st to Saa
S
technol
ogy
, and it can h
e
lp
the use
r
to a
c
hieve o
p
e
r
at
ion se
rvice of prod
uct
on t
he Intern
et. at the same ti
me, the platform
as a develo
p
m
ent tool can
be used a
s
a powe
r
ful
su
ppleme
n
t of core ele
c
tro
n
ic devices, hig
h
-
end ge
neral chip and b
a
si
c softwa
r
e pro
duct
s
.
Ackn
o
w
l
e
dg
ement
This work is
suppo
rted by Natural Sci
e
n
c
e Fou
ndatio
n of China (No. 61262
049
), the
Re
sea
r
ch Pla
n
of departm
ent of education of
Jian
gxi Province (G
JJ12
197
). The
authors are
grateful for th
e anonymo
us reviewe
r
s wh
o made con
s
tructive
comm
ents.
Referen
ces
[1]
A.losup et al.
”Performance Analy
s
is of Clo
ud Comput
ing Servic
es for Many
-T
asks Scientific
Comp
uting”.
IEEE Trans. Parallel an
d Distributed System
s
. 2011; 22(
6): 931-9
45.
[2]
C.Vecchi
o
la,
S.Pande
y, a
n
d
R.Bu
yya. ”H
igh-P
e
rforman
c
e Clo
ud C
o
mputin
g:
A Vie
w
of Scie
ntifi
c
Appl
icatio
ns”. Proc, 10th Int’l
S
y
mp. Perv
asi
v
e
S
y
stems, Al
gorithms, an
d
Net
w
orks (ISP
AN 09).
IEEE
CS
. 2009: 4-
16
.
[3]
J Li et al. “eScienc
e in the Cl
oud: A Modis Sate
ll
ite Data Repr
ojecti
on a
nd Red
u
ctio
n Pipe
lin
e in t
h
e
Windo
w
s
Az
ur
e Platform”. Proc.2 010 Intl S
y
mp.
Par
a
ll
el
& Distribute
d
Processi
ng(IP
DPS 10).
IEEE
CS
. 2010: 1-
10
.
[4]
J Dean and S Ghema
w
at. ”Map Red
u
ce: Simplifi
ed Dat
a
Processin
g
on Larg
e
Clus
t
ers”. Comm.
ACM
. 2008: 5
1
(
1): 107-1
13.
[5]
Jeffre
y
De
an
and Sa
nja
y
G
hema
w
a
t. Mapred
uce: Simp
lified d
a
ta pro
c
essin
g
on lar
ge clusters.
Co
mmun.A
C
M
. 2008; 51(
1): 107–
11
3.
[6]
I F
o
ster, Y
Z
hao, I Raicu, an
d S Lu. “Clo
ud
C
o
mputi
ng a
nd
Grid Comp
utin
g 360-
de
gree c
o
mpar
ed”.
in
Grid Co
mp
utin
g Enviro
n
m
ent
s W
o
rkshop
. 2
008; 1–
10.
[7]
G DeCan
d
ia,
D Hastoru
n
, M Jampan
i, G Kakul
apati, A La
kshman, A Pilc
hi
n, S Sivasu
b
r
amani
an, P
Vossha
ll, an
d W
Vogels. “D
ynamo: Amazo
n
’s hi
ghl
y av
ai
labl
e ke
y-va
lu
e store”. in S
y
mposi
u
m
o
n
Operatin
g S
y
st
ems Princi
ples.
ACM
. 2007: 2
05–
22
0.
[8]
P W
a
tson, P Lord, F Gibson, P Periore
llis, a
nd G Pitsilis. Cl
oud C
o
mputi
n
g
for e-Science
w
i
t
h
CARMEN.In 2
nd Iberi
an
Grid
Infrastructure Confer
ence Pr
ocee
din
g
s
. 200
8; 3–14.
[9]
P Mell and T
Grance. ”Effectivel
y
an
d Secure
l
y
Usi
n
g the Cloud Comp
uting ar
adi
gm (v0.25)”
.
Presentati
on,
US Nat
’
l Inst. Standar
ds an
d T
e
chn
o
lo
gy
. 200
9.
Evaluation Warning : The document was created with Spire.PDF for Python.
TELKOM
NIKA
ISSN:
2302-4
046
A SaaS Deve
lopm
ent Platform
base
d
on
Cloud
Com
puting (Qin
gpe
ng Zeng
)
1651
[10]
M Ha
ynie. “Ent
erprise cl
ou
d services:
Deriv
i
ng bus
iness v
a
lu
e from Clou
d
Comp
uting”.
Micro Focus,
Tech. Rep
. 20
09.
[11]
M Armbrust, A
F
o
x, R Griffith,
A Joseph, R Katz
, A Kon
w
i
n
s
k
i, G Lee, D Patterson, A Rabkin, I Stoica,
and M Z
a
h
a
ria
.
“
Above the
Clou
d
s: A Ber
k
eley
vi
ew
of Clou
d
Co
mputi
n
g
”. Univ
ersit
y
of Califor
nia,
Berkel
e
y
, T
e
ch. Rep. 200
9.
[12]
Amazon EC2. http
://aw
s
.amazon.com/ec2.
[13]
Hadoop. http
://hadoop.apac
he.org.
[14]
Open Cl
ou
d Consorti
um. h
ttp://
w
w
w
.
op
enc
lo
udco
n
sortium.
org.
Evaluation Warning : The document was created with Spire.PDF for Python.