Sei sulla pagina 1di 54

1esLlng MeLhodologles 1esLlng MeLhodologles

1oplcs 1oplcs
Cb[ecLlve
SofLware LesLlng WhaL? Why? Who?
8ole of 1esL aL MlcrosofL
Challenges
lew baslc LesLlng quesLlon's
lnLroducLlon Lo baslc LesLlng prlnclples
1esL Llfe cycle
1esLlng Approaches
1esLlng levels
1esLlng 1ypes/ MeLhodologles
SofLware 1esL ArLlfacLs
1esL lan
1esL Case
8ug Llfe cycle
8esL racLlces ln 1esLlng

Cb[ecLlve Cb[ecLlve
undersLand whaL why and who of LesLlng?
Share undersLandlng of LesLs role and challenges aL MlcrosofL
know abouL Lhe 1esL llfe cycles
undersLand Lhe dlfferenL LesLlng approach's/ levels/ Lypes /
meLhodologles aL MlcrosofL
Learn how Lo creaLe a LesL plan deslgn LesL cases and bug llfe
cycle
And mosL lmporLanL become all seL Lo LesL MSn pro[ecLs

WhaL ls SofLware 1esLlng? WhaL ls SofLware 1esLlng?


1estinq is 4N 4k1 Ol cON51kuc1lv ucON51kuc1lON
An lLeraLlve process conslsLlng of seL of acLlvlLles such as WrlLlng 1esL lans
ueveloplng LesL cases 8unnlng LesL cases and 8eporLlng uefecLs
AL MlcrosofL LesLlng ls Lhe acL or process of applylng a LhoughL or ldea as a means Lo
analyze or dlagnose
Many ln Lhe lndusLry argue Lhe role of LesLlng ls Lo flnd bugs Lo prove errors eLc 1hls
approach mlsuses Lhe aLLempL of quallLy conLrol process Lo LesL quallLy ln"
ln order Lo lmprove overall quallLy we need Lo
Move fity stref
9revent defects

WhaL ls SofLware 1esLlng? WhaL ls SofLware 1esLlng?


9rifry qofs of the enqineerinq exceence initiftives for test?
uefecL prevenLlon by engaglng LesLlng ln every aspecL of Lhe producL cycle
urlvlng quallLy upsLream
CusLomer focus
CuallLy measuremenL

ow do you deflne 8ug/CuallLy? ow do you deflne 8ug/CuallLy?


8ug
A devlaLlon from cusLomer expecLaLlon
An error ln code or loglc
MlslnLerpreLaLlon of speclflcaLlon
;ua||ty
Level of Lxcellence
uegree Lo whlch a producL fulflls cusLomer requlremenLs
A good quallLy sofLware funcLlons as expecLed ls easy
Lo use sLable and bug free

Why Lo LesL? Why Lo LesL?


Jy?
umans LesL Lhlngs all Lhe Llme nobody would buy anyLhlng expenslve or complex
wlLhouL some klnd of evaluaLlon
ost fry is chef
lL lncreases exponenLlally as Lhe pro[ecL develops
1 Spec revlew
uurlng developmenL
uurlng LesLlng
uurlng locallzaLlon
osL producLlon
;ua||ty 1he earller Lhe lssues found Lhe greaLer chance of lL geLLlng flxed Lo release
a quallLy producL
va|uat|on lf Lhe 8CuuC1 maLches Lhe expecLaLlons for a glven Llme

Who performs 1esLlng? Who performs 1esLlng?


Jo tests?
veryone!!
1esLer
uevelopers
rogram Managers
Ceneral/upper ManagemenL

8ole of 1esL aL MlcrosofL 8ole of 1esL aL MlcrosofL


vocate
for qua||ty
Unerstan
process not [ust
prouct
8e te
vo|ce of
customer
4 #s aut|ous
ur|ous r|t|ca|
an onv|nc|ng
Sou| o
Ia|s|f|cat|on
Va||at|on rror
prevent|on k|sk
ana|ys|s
10
Challenges Challenges
ggress|ve
sceu|e
omp|ex|ty
Irequent
re|eases
11
lew 8aslc 1esLlng CuesLlon's lew 8aslc 1esLlng CuesLlon's
Why s/w has bugs?
uoes every pro[ecL need LesLlng?
WhaL lf Lhere ls noL enough Llme?
When Lo sLop LesLlng?
WhaL ls lmporLance of documenLaLlon
ln LesLlng?
WhaL ls blg deal abouL requlremenLs?
1
lnLroducLlon Lo 8aslc 1esLlng rlnclples lnLroducLlon Lo 8aslc 1esLlng rlnclples
1 1esLlng shows presence of defecLs
LxhausLlve LesLlng ls lmposslble
Larly LesLlng
uefecL clusLerlng
esLlclde paradox
1esLlng ls conLexL dependenL
Absenceoferrors fallacy
1
1esL Llfe Cycle 1esL Llfe Cycle
1
1esLlng Approaches 1esLlng Approaches
8aslc approach's Lo LesL sofLware
8lack 8ox
Crey 8ox
WhlLe 8ox
8|ack 8ox
Jat?
1esLer does noL (or should noL) have access Lo Lhe source code
1esLlng from user perspecLlve
1esLs developed Lo verlfy funcLlonallLy of sofLware
Jy?
1o cover Lhe funcLlonallLles speclfled ln Lhe funcLlonal speclflcaLlons
1o uncover Lhe excepLlonal usage scenarlos
xamp|e ? ln a Search englne one of Lhe black box LesL case would be verlfylng wheLher Lhe
Search resulLs are feLched on cllck of Search buLLon or noL
1
1esLlng Approaches 1esLlng Approaches
rey 8ox
Jat?
1esLer has access Lo lnLernal daLa sLrucLures and algorlLhms for deslgnlng Lhe LesL
cases
1esLlng ls done aL Lhe user or blackbox level
Jy?
8eLLer LesL coverage
uefecLs relaLed Lo bad deslgn or bad lmplemenLaLlon of Lhe sysLem are uncovered
xamp|e ? Lmall valldaLlon uslng scrlpL ln webslLe LesLlng we can have a LesL case on cllenLs
where [ava scrlpLs are dlsabled lf we know abouL Lhe lmplemenLaLlon deLalls
J|te 8ox
Jat?
8ased on knowledge of lnLernal loglc of a parLlcular appllcaLlon code
1esLer develops LesLs Lo verlfy funcLlons ln code
1
1esLlng Approaches 1esLlng Approaches
Jy?
Lxerclses
All lndependenL paLhs/ loglcal declslons/ loops (aL Lhelr boundarles and wlLhln
Lhelr operaLlonal bounds)
lnLernal daLa sLrucLures Lo ensure Lhelr valldlLy
valldaLe code flxes prlor Lo check ln by applylng dpks
WrlLe beLLer bugs lncludlng debugglng lnformaLlon saves uLv lnvesLlgaLlon Llme
and speeds bug Lurn around
xamp|e? Loop 1esLlng
1
1esLlng Levels 1esLlng Levels
System TestIng
ntegratIon TestIng
Component TestIng
UnIt TestIng
1
1esLlng Levels 1esLlng Levels
&39%0893
1esLlng performed on lndlvldual parLs of code ( SubrouLlne funcLlons)
WhlLe box LesLlng
1yplcally performed by developer or SuL1
42543039%0893
1esLlng lnLeracLlon beLween mulLlple unlLs" and daLa ob[ecLs
WhlLe /grey or black box LesLlng
1yplcally performed by SuL1 /S1L
3907,943%0893
1esLlng Lhe lnLeracLlon of lnLegraLed componenLs (already LesLed)
WhlLe /grey or black box LesLlng
1yplcally performed by SuL1 /S1L
1
1esLlng Levels 1esLlng Levels
$8902%0893
1esLlng Lhe robusLness of Lhe sysLem and lLs lnLeracLlons wlLh exLernal sysLems
1esLlng Lhe usablllLy and funcLlonallLy of Lhe sofLware ln a real world"
envlronmenL
8lack box LesLlng
1yplcally performed by SuL1/ S1L
st inforftion u14lLu 1k4lNlN65 klk 1O 991s
hLLp//hcllnslghL
0
1esLlng 1ypes/MeLhodologles/LxecuLlon rocess 1esLlng 1ypes/MeLhodologles/LxecuLlon rocess
4 brief overview of efch qiven in the sides fhefd
u14lLu 1k4lNlN65 klk 1O 991s
hLLp//hcllnslghL
@ypes @ypes Metoo|og|es xecut|on rocess
ul erformance Model 8ased Smoke
usablllLy MonlLorlng Code Coverage 8v1
LocallzaLlon lall Cver laulL ln[ecLlon lull 1esL ass
uaLabase Compllance Aglle Ad hoc
Al Lqulvalence
parLlLlonlng
8egresslon
Web Servlces 8oundary value
Analysls
Culck 1esL ass
lnsLallaLlon AccepLance
ueploymenL
1
1esLlng 1ypes/MeLhodologles 1esLlng 1ypes/MeLhodologles
User Interface (UI)
ey to fccetfnce of end rodct!!
Checklng wheLher Lhe ul look and feel ls pleaslng or noL? WheLher Lhe ul elemenLs
are dlsplayed properly or noL? WheLher Lhey are allgned properly or noL? Are Lhey ln
correcL order or noL and so on
@oo|s lL uev Loolbar llrebug eLc
xamp|e? 1ypography (8ed llne spec) (Lg @aoma |s te pr|mary on||ne font for MSN)
Usab|||ty
ase of useDone before feature]prouct |aunc
ln a usablllLy LesL
A usablllLy lab ls seL up
8epresenLaLlve users Lry Lo do Lyplcal Lasks wlLh Lhe producL
Cbservers lncludlng Lhe developmenL sLaff waLch llsLen and Lake noLes
Measures how well LesL sub[ecLs (users) respond ln four areas efflclency
accuracy recall and emoLlonal response

1esLlng 1ypes/MeLhodologles 1esLlng 1ypes/MeLhodologles


Moe| 8ase
An obsttoctloo or slmpllfled tepteseototloo of Lhe sysLem from a parLlcular
perspecLlve
omples llow CharLs ulu's L8 ulagrams SLaLe ulagrams use Cases
AcLlvlLy ulagrams eLc
@pes SLaLeless and sLaLeful
Jy?
Larly bug deLecLlon (many are found ln spec phase)
1esLlng beglns wlLh enduser scenarlos and flnds posslble or mlsslng LranslLlon
scenarlos
uraws effecLlve parLlclpaLlon of uev/M for 8evlews
Lasy Lo updaLe Lhe model for enhancemenLs and fuLure verslons of producL
@oo|s SpecLxplorer lC1 kokomo Ceno MuL 8A11
Model urlven 1esL ueslgn Model urlven 1esL ueslgn

softwure
urtifuct
modeI /
structure
test
requirements
refined
requirements /
test specs
input
vuIues
test
cuses
test
scripts
test
resuIts
puss /
fuiI
--~~~
~:~^~
--
.-,~
~:~^~
--
Model urlven 1esL ueslgn Model urlven 1esL ueslgn SLeps SLeps

softwure
urtifuct
modeI /
structure
test
requirements
refined
requirements /
test specs
input
vuIues
test
cuses
test
scripts
test
resuIts
puss /
fuiI
--~~~
~:~^~
--
.-,~
~:~^~
--
anaIysis
criterion refine
generate
prefix
postfix
expected
automate
execute
evaIuate
Model urlven 1esL ueslgn Model urlven 1esL ueslgn AcLlvlLles AcLlvlLles

softwure
urtifuct
modeI /
structure
test
requirements
refined
requirements /
test specs
input
vuIues
test
cuses
test
scripts
test
resuIts
puss /
fuiI
--~~~
~:~^~
--
.-,~
~:~^~
--
%est Design %est Design
%est %est
Execution Execution
%est %est
EvaIuation EvaIuation
[~.-.:g cu ~-r~cr.c: |e.e| ~|e-
re-r Je-.g: |__| e~-.e

1esLlng 1ypes/MeLhodologles 1esLlng 1ypes/MeLhodologles


pp||cat|on rogramm|ng Interface (I)
A collecLlon of sofLware funcLlons and procedures called Al calls LhaL can
be execuLed by oLher sofLware appllcaLlons
ulfferenL from oLher LesLlng Lypes as Cul ls rarely lnvolved ln Al LesLlng
@oo|s WCl 1esLCllenL
Jeb Serv|ces
Web Servlces exLend Lhe World Wlde Web lnfrasLrucLure Lo provlde Lhe means
for sofLware Lo connecL Lo oLher sofLware appllcaLlons
AppllcaLlons access Web servlces vla Web proLocols and
daLa formaLs such as 11 xML and SCA
@oo|s llddler 11 WaLch WS 1ool

1esLlng 1ypes/MeLhodologles 1esLlng 1ypes/MeLhodologles


oca||zat|on
4 ste fhefd to word kefdy softwfre {6obfited) !!
1esLlng Lo check how well Lhe bulld has been LranslaLed lnLo a parLlcular LargeL
language
@oo|s LocSLudlo
xamp|e? AssessmenL of culLural approprlaLeness Checklng for pollLlcally senslLlve conLenL eLc
ommon prob|ems CuesLlon marks (????) appearlng lnsLead of dlsplayed LexL 8andom characLers
lnsLead of readable LexL eLc
Iau|t In[ect|on
1o deveo kobst softwfre!!
A Lechnlque Lo lmprove Lhe LesL coverage by lnLroduclng faulLs Lo LesL code paLhs (
malnly error handllng code paLhs)
teote bpotbeslzeJ ettots by Addlng/ changlng/ modlfylng Lhe
code under analysls
@pes Complle 1lme and 8unLlme
xamp|e? omp||e t|me changlng a+1 Lo a1 kunt|me CorrupLlon of memory space

1esLlng 1ypes/MeLhodologles 1esLlng 1ypes/MeLhodologles


Database
1esLlng Lo make sure Lhe code really ls readlng and wrlLlng Lhe rlghL daLa Lo
Lhe daLabase
ueLermlnes how well Lhe daLabase meeLs requlremenLs
lmprove overall 1esL sLraLegy and producL's quallLy
ocos Ateos uaLa valldlLy uaLa lnLegrlLy 1esLlng of
rocedure 1rlggers and luncLlons Checklng for any uaLa Loss
erformance MonlLorlng ScalablllLy SecurlLy of uaLa eLc
@oo|s dbnunlL SCL rofller

1esLlng 1ypes/MeLhodologles 1esLlng 1ypes/MeLhodologles


oe overage
4 dynfic whitebox test fssessent technie
Shows whaL has been and has noL been execuLed (LesLed)
8equlres access Lo source code
lmprove LesL sulLe ( ldenLlfles weak area's)
uurlng unlL LesLlng and SysLem LesLlng
@oo|s Magellan
erformance @est|ng
rocess of measurlng how fasL and rellable Lhe sysLem ls funcLlonlng wlLh
varylng level of user load
ow ls performance ls measured
CllenL slde measure how Lhe user experlences Lhe performance of Lhe
web slLe lrom user erspecLlve
0
1esLlng 1ypes/MeLhodologles 1esLlng 1ypes/MeLhodologles
Server slde measure how Lhe servers are handllng Lhe requesLs from Lhe cllenL lrom
CperaLlons erspecLlve
@oo|s erfMon erfAnalyzer erf 1esL Lhrough vS1S
Insta||at|on
lirst interfction of ser with or rodct
ocos oteos lnsLallaLlon and un lnsLallaLlon
@pes SllenL unaLLended llaL CLhers (Self lnsLallaLlon Clean lnsLallaLlon )
Dep|oyment
1ypes of ueploymenL Slngle box and MulLl box
5inqe box ln Lhls Lype Lhe 8L lL and Lhe daLabase reslde ln Lhe same machlne and
Lhe cllenLs connecL from oLher machlnes
1
1esLlng 1ypes/MeLhodologles 1esLlng 1ypes/MeLhodologles
,ti box ln Lhls Lype Lhe 8L and lL are spllL and made Lo run from dlfferenL machlnes
1hls ls malnly employed where you need Lo have very hlgh performance 1he lL handles
mosL of Lhe 8ead requesLs from Lhe cllenL and forwards only Lhe WrlLe requesLs Lo Lhe
8L provldlng hlgh performance
uto Dep|oyment A meLhod Lo qulckly deploy an appllcaLlon Lo a sLandalone server
(Admln Server) for evaluaLlon or LesLlng
8ecommended only ln a slngleserver developmenL envlronmenL
Jy Dep|oyment @est|ng?
ldenLlfy lssues proacLlvely LhaL mlghL occur ln Lhe deploymenL slLe
@oo|s CcLopus AuLoplloL

1esLlng 1ypes/MeLhodologles 1esLlng 1ypes/MeLhodologles


omp||ance @est|ng
Compllance accepLance LesLlng or regulaLlon accepLance LesLlng ls
performed agalnsL Lhe regulaLlons whlch musL be adhered Lo such as
governmenLal legal or safeLy regulaLlons
ocos Ateos SecurlLy rlvacy CompaLlblllLy and AccesslblllLy
Mon|tor|ng @est|ng
Lnsures LhaL Lhe sysLem whlch ls geLLlng deployed provldes enough lnformaLlon Lo
Lhe operaLlons personnel Lo assure LhaL Lhe sysLem ls runnlng properly
8equlred for Webbased appllcaLlons and mlssloncrlLlcal enLerprlse appllcaLlons LhaL
musL Lyplcally provlde roundLheclock servlces
CeLs necessary fallure avallablllLy and Lrend analysls daLa of Lhe sysLem
@oo|s x1S neLMon unS Look up

1esLlng 1ypes/MeLhodologles 1esLlng 1ypes/MeLhodologles


Ia|| Cver
4 ste to fvoid 9rodction Otfqe
lor example ln a web envlronmenL fallover LesLlng deLermlnes whaL
wlll happen lf mulLlple web servers are belng used under peak anLlclpaLed
load and one of Lhem dles
uoes Lhe load balancer reacL qulckly enough?
Can Lhe oLher web servers handle Lhe sudden dumplng of exLra load?
8u|| Ver|f|cat|on @est (8V@)
Crouplng of crlLlcal 0 LesLs cases Lo cover Lhe ma[orlLy of funcLlonallLy and accepL Lhe
bulld for furLher LesLlng
1yplcally run on dally bullds and lf Lhe 8v1 falls Lhe bulld ls re[ecLed and a new bulld ls
released afLer Lhe flxes are done
Saves Lhe efforLs of a LesL Leam Lo seLup and LesL a bulld when ma[or funcLlonallLy ls
broken
1yplcally 8v1's should noL lasL for more Lhan 0 mlnuLes
@oo|s W11 L1CM ( auLomaLed execuLlon of 8v1's)

1esLlng 1ypes/MeLhodologles 1esLlng 1ypes/MeLhodologles


oc @est|ng
1esLlng ln an unplanned unsLrucLured manner
ConslsLs of scenarlo's oLher Lhan 0 1
coverlng Lhe funcLlonallLy
Smoke @est|ng
lnlLlal LesL on sofLware afLer lLs creaLlon or afLer a bug ls flxed
verlfles hlgh level funcLlonallLy before code ls checked lnLo bulld

1esLlng 1ypes/MeLhodologles 1esLlng 1ypes/MeLhodologles


kegress|on @est|ng
8erunnlng exlsLlng LesLs agalnsL Lhe modlfled code Lo deLermlne wheLher Lhe
changes break anyLhlng LhaL worked prlor Lo Lhe change
CreaLe and run new LesL cases lf requlred
9tlmot ooslJetotloo AdequaLe coverage wlLhouL wasLlng Llme
cceptance @est|ng
8lackbox LesLlng performed on a sysLem prlor Lo lLs dellvery
A lormal LesL conducLed Lo deLermlne wheLher or noL a sysLem saLlsfles lLs
accepLance crlLerla
Lnables cusLomer Lo deLermlne wheLher or noL Lo accepL Lhe sysLem

1esLlng 1ypes/MeLhodologles 1esLlng 1ypes/MeLhodologles


Iu|| @est ass
ConslsLs of execuLlng all Lhe 0 1 level LesL cases coverlng breadLh and depLh
aspecLs agalnsL Lhe feaLures dellvered ln Lhe bulld
oa|s
Shlp a Llmely and quallLy producL
CeL full coverage for all Lhe feaLures
Make Lhe correcL Lradeoffs for Lhe conflguraLlon maLrlx
undersLand prlorlLles and mulLlLask
1lme Lo meeL or beaL deadllnes

SofLware 1esL ArLlfacLs SofLware 1esL ArLlfacLs


1esL case/ scrlpL
1esL daLa
1esL SulLe
1esL lan
1esL arness
1esL ueslgn 1esL ueslgn
1esL log
1esL Summary

1esL lan WhaL? Why? Who? 1esL lan WhaL? Why? Who?
Jat
MosL slgnlflcanL LesLlng documenL!!
8eflecLs enLlre pro[ecL LesLlng schedule and approach
CapLure lnformaLlon from varlous sLakeholders ( uev M
cusLomer) of Lhe pro[ecL
Jy?
LsLlmaLe Lhe requlred efforLs
undersLand why and how of producL valldaLlon
Clarlfy Lhe ob[ecLlves scope focus and approach of sofLware LesLlng
ConfronL challenges and crysLalllze Lhlnklng
Jo?
usually creaLed by 1esL Manager 1esL Lead or senlor LesLers ln Lhe Leam

1esL lan CharacLerlsLlcs/1lps 1esL lan CharacLerlsLlcs/1lps


aracter|st|cs of a oo @est |an
Clear purpose
Accesslble
MalnLalnable
usable
Well organlzed
CompleLe
@|ps for wr|t|ng an effect|ve @est p|an
use drafLs
use bralnsLormlng sesslons
old formal revlew
keep LesLlng focus
Manage granularlLy
know your audlence's expecLaLlons
0
1esL lan Ma[or Area's 1esL lan Ma[or Area's
Introuct|on urpose Summary 8evlslon lsLory
rerequ|s|te#s uependencles AssumpLlons
;ua||ty Cb[ect|ve Spec/1esL arLlfacL 8evlews
Scope ln scope ouL of scope
k|sk Area's 1esL 1eam uevelopmenL
MarkeL CusLomer lnclude mlLlgaLlon
sLraLegy as well
@est pproac]Strategy lrameworks/1ypes of 1esLlng/ Speclflc
sLraLegy relaLed Lo pro[ecL
@est Specs Scenarlo's LesL cases llnks
; ro[ect Management 1esL Case managemenL
1esL asses
1esL AccepLance /
8elease/LxlL CrlLerla
SLaLus 8eporLlng
1
1esL lan Ma[or Area's 1esL lan Ma[or Area's
nv|ronment SofLware ardware deLalls
@est De||verab|e#s 1esL lan scenarlo's LesL cases resulL
logs execuLlon summary sLaLus
reporLs bug reporLs
Sceu|e]kesources
ppen|x

1esL lan 1hlngs Lo Conslder 1esL lan 1hlngs Lo Conslder


Ask Ooestloos you may flnd yourself reenglneerlng your LesL seLup before LesLlng even beglns!
2 Aotlclpote where your producL may lAlL and uLvLLC S18A1LClLS Lo address poLenLlal
problems
lgh level LesL plan should address Lhe overall requlremenLs and supporLlng LesL plans should
address Lhe deslgn deLalls of subsysLems and componenLs
And mosL lmporLanL Lhe Llme you spend 8LSLA8ClnC Anu uLSlCnlnC
your LesL plan before you sLarL LesLlng wlll SAvL ?Cu MCnL?
Sample 1esL lan 1emplaLes Sample 1esL lan 1emplaLes

rea Samp|e @est |an


MasLer 1esL lan
leaLure/SubfeaLure 1esL lan
erformance 1esL lan
SecurlLy 1esL lan
ul 1esL lan
Naster Test Plan
Feature test Plan
Performance Test
Plan
Security Test Plan
U! Test Plan
Sample 1esL lan 1emplaLes Sample 1esL lan 1emplaLes

rea Samp|e @est |an


1esL AuLomaLlon lan
8elease 1esL lan
uaLabase 1esL lan
8rowser CompaLlblllLy
AccepLance 1esL lan
utomation Test
Plan
Release Test Plan
Database Test Plan
Compatibility Test
Plan
UT Test Plan
1esL Case WhaL? ComponenLs? 1ypes? 1esL Case WhaL? ComponenLs? 1ypes?
Jat?
A quallLy assessmenL acLlvlLy LhaL relaLes a LesL lLem and lnpuL Lo expecLed ouLpuL
1esLer's one of Lhe lmporLanL uellverable
aracter|st|cs of a oo @est ase
8evlewable
CompleLe
8eusable
LfflclenL
CLhers

1esL Case WhaL? ComponenLs? 1ypes? 1esL Case WhaL? ComponenLs? 1ypes?
omponents of an ffect|ve @est ase
lnpuL and ouLpuL
1esL case ldenLlfler
1esL lLems
LnvlronmenLal needs
lnLer case dependencles
CLher 8aslc deLalls
@est ase r|or|t|es? 0 1
0 key Lo Lhe producL/feaLure worklng aL all 8v1 level
1 8aslc funcLlonallLy for key scenarlos Smoke level LesLs AccepLance
level LesLs
2 8eyond baslc funcLlonallLy buL sLlll a falrly common user experlence A
small percenLage may make Lhe accepLance level LesL case mosL are
comprehenslve

1esL Case WhaL? ComponenLs? 1ypes? 1esL Case WhaL? ComponenLs? 1ypes?
AnoLher Lermlnology whlch ls used aL MlcrosofL ls
@est ase eve|s? L0 L1 L L
0 (AccepLance 1esLlng aka 8v1) L0's passlng would lndlcaLe a valuable
bulld Lo use for exLended LesLlng
1 (valldlLy 1esLlng Lhe pro[ecL/feaLure works) L1's passlng would lndlcaLe
valuable bulld for lnLernal cusLomers parLners eLc
2 (lnvalld/Lrror/8oundary 1esLlng) L's passlng would lndlcaLe valuable
and robusL bulld for lnLernal cusLomers parLners eLc 8ulld behaves
accepLably when used lncorrecLly
(LxLreme 8obusLness 1esLlng) roducL may sLlll shlp even lf a problem
was found wlLh Lhls Lype of LesL dependlng on Lhe scenarlo Level LesLlng
ls very lmporLanL for undersLandlng Lhe Lrue quallLy of Lhe producL

ueslgnlng 1esL Cases ueslgnlng 1esL Cases


ommon @ypes
valldaLlon vs lalslflcaLlon
SLaLlc LesL cases
uynamlc LesL cases
luncLlonal LesL cases
lnLegraLlon LesL cases

8ug Llfe Cycle 8ug Llfe Cycle


lf you are a sofLware LesLer or a CA englneer Lhen you musL be Lhlnklng every
mlnuLe Lo loJ a boq ln an appllcaLlon
A bug occurs due Lo a devlaLlon from cusLomer expecLaLlon an error ln code or loglc or
mlslnLerpreLaLlon of speclflcaLlon eLc
@ypes of 8ugs
usablllLy
uocumenLaLlon
LnvlronmenLal
luncLlonal
lnLerface
erformance
Spec 8evlew
ux
CloballzaLlon
Load/SLress
SecurlLy

8ug Llfe Cycle 8ug Llfe Cycle


8ug |fe yc|e
8ug ls Cpened AcLlve
8ug ls @r|age ( A meLhod of raLlonlng llmlLed developmenL and LesL resources Lo
address Lhe mosL lmporLanL lssues by prlorlLlzlng bugs asslgnlng Lo correcL area/dev
deflnlng a mllesLone eLc)
8ug ls 8esolved ( llxed/ noL repro/Won'L flx /ueslgn eLc)
8ug ls verlfled as flxed and closed by 1esL
@oo| roducL SLudlo
0
$oftware undergoing the same repetitive
tests eventuaIIy buiIds up resistance to
them.
8ug Llfe Cycle 8ug Llfe Cycle
1
8esL 8esL racLlses racLlses
1 SLarL LesLlng acLlvlLles aL Lhe beg|nn|ng of SofLware developmenL
undersLand Lhe requlremenLs ln ept
updaLe Lhe LesL cases lmmedlaLely once Lhe requlremenL change
MalnLaln a LraceablllLy maLrlx
1esL cases should be clear effecLlve and descrlpLlve
lannlng and creaLlng LesL plan documenL ls essenLlal even for
small shorLLerm pro[ecLs
erform retest and small regress|on test whenever a reporLed
bug ls flxed
uo adhoc ln addlLlon Lo LesL case execuLlon

Mow - I oIreody
don'f pIow os good
os I know how...
They're feoching o new woy
of pIowing over of fhe 0ronge
fonighf - you going7
"Knowing is not enough, we must apply. Willing is not enough, "Knowing is not enough, we must apply. Willing is not enough, we we
must do must do. Goethe . Goethe
CuesLlon's CuesLlon's

Potrebbero piacerti anche