Professional Documents
Culture Documents
OBIEE Interview Questions and Answers
OBIEE Interview Questions and Answers
Define pipeline. Did you use it in your projects? o Yes, pipelines are the stages in a particular transaction. Assessment, finance etc. How do you create filter on repository? o Where condition on content tab. How do you work in a multi user environment? !at are t!e steps? o Create a shared directory on the network for Multi-user De elopment !M"D#. o $pen the rpd to use in M"D. %rom &ools-'$ptions, setup the M"D directory to point to the abo e directory. o Define pro(ects within the rpd to allow multiple users to de elop within their sub(ect area or %acts. o )a e and mo e the rpd to the shared directory setup in point *. o When users work in the M"D mode, they open the admin tool and start with o M"D -'Checkout to checkout the pro(ect they need to work on !not use the %ile open as you would usually do#. o After completely the de elopment, user checkin the changes back to the network and merge the changes. !ere are passwords for userid? "dap#e$ternal ta%le aut!entication stored respectively? o passwords for userid are in siebel analytics ser er repository +dap authentication in +dap ser er e,ternal database in a table in e,ternal database &an you %ypass 'ie%el analytics server security ?if so !ow? o yes you can by-pass by setting authententication type in -.)C$-%/0 file in the security section as1authentication2type3bypass2n4s.instanceconfig.,ml and n4sconfig.ini are the 5 places Where can you add new groups and set permissions6 o you can add groups by going to manage'security'add new groups' You can gi e permissions to a group for 4uery limitation and filter conditions. w!at are t!e t!in(s you can do in t!e B)) layer? o Aggregation na igation, le el base metrics, time series wi7ard, create new logical column, comple, (oin. w!at is *a((ed !ierarc!y? and !ow do u mana(e it o 8agged 9ierarchy is one of the different kinds of hierarchy. o A hierarchy in which each le el has a consistent meaning, but the branches ha e inconsistent depths because at least one member attribute in a branch le el is unpopulated. A ragged hierarchy can represent a geographic hierarchy in which the meaning of each le el such as city or country is used consistently, but the depth of the hierarchy aries. o %or e,ample, a geographic hierarchy that has Continent, Country, :ro ince;)tate, and City le els defined. $ne branch has -orth America as the Continent, "nited )tates as the Country, California as the :ro ince or )tate, and )an %rancisco as the City. 9owe er, the hierarchy becomes ragged when one member does not ha e an entry at all of the le els. %or e,ample, another branch has <urope as the Continent, 0reece as the Country, and Athens as the City, but has no entry for the :ro ince or )tate le el because this le el is not applicable to 0reece for the business model in this e,ample. /n this e,ample, the 0reece and "nited )tates branches descend
to different depths, creating a ragged hierarchy. !at is t!e difference %etween 'in(le "o(ical +a%le 'ource and )ultiple "o(ical +a%le 'ources? o /f a logical table in =MM layer has only one &able as the source table then it is )ingle +&). o /f the logical table in =MM layer has more than one table as the sources to it then it is called Multiple +&). o <,1 "sually %act table has Multiple +&)>, for which sources will be coming from different :hysical tables. Can you let me know how many aggregate tables you ha e in your pro(ect6 $n what basis ha e you created them6 o As per resume (ustification document How do you %rin(,relate t!e a((re(ate ta%les into t!e 'ie%el analytics "o(ical layer? o $ne way of bringing the Aggregate &ables into the =MM layer is by bringing them as +ogical &able sources for the corresponding %act table. o &his is done by dragging and dropping the aggregate table into the corresponding fact table. After doing that establish the column mappings and the set the aggregation le els. How do you know w!ic! report is !ittin( w!ic! ta%le# eit!er t!e fact ta%le or t!e a((re(ate ta%le? o After running the report, go to Administration tab and go to click on Manage )essions. &here you can find the 4ueries that are run and in the ?iew +og option in the )ession Management you can find which report is hitting which table. 'uppose I !ave report w!ic! is runnin( for a%out - minutes typically. !at is t!e first step you take to improve t!e performance of t!e .uery? o %ind the s4l 4uery of the report in Admin-'manage )ession-' run the s4l 4uery on toad -'read the e,plain plan output -'modify the ).+ based on the e,plain plan output 'uppose you !ave a report w!ic! !as t!e option of runnin( on a((re(ate ta%le. How does t!e tool know to !it t!e A((re(ate ta%le and for t!at w!at t!e steps you follow to confi(ure t!em? o <,plain the process of Aggregate na igation 9a e you heard of /mplicit %acts6 /f, so what are they6 o An implicit fact column is a column that will be added to a 4uery when it contains columns from two or more dimension tables and no measures. You will not see the column in the results. /t is used to specify a default (oin path between dimension tables when there are se eral possible alternati es. o %or e,ample, there might be many star schemas in the database that ha e the Campaign dimension and the Customer dimension, such as the following stars1 Campaign 9istory star. )tores customers targeted in campaign. Campaign 8esponse star. )tores customer responses to a campaign. $rder star. )tores customers who placed orders as a result of a campaign. /n this e,ample, because Campaign and Customer information might appear in many segmentation catalogs, users selecting to count customers from the targeted campaigns catalog would be e,pecting to count customers that ha e been targeted in specific campaigns. &o make sure that the (oin relationship between Customers and Campaigns is through the campaign history fact table, a campaign history implicit fact needs to be specified in Campaign
9istory segmentation catalog. &he following guidelines should be followed in creating segmentation catalogs1 <ach segmentation catalog should be created so that all columns come from only one physical star. =ecause the Marketing module user interface has special features that allow users to specify their aggregations, le el-based measures typically should not be e,posed to segmentation users in a segmentation catalog. !at is a((re(ate navi(ation? How do you confi(ure t!e A((re(ate ta%les in 'ie%el Analytics? o Aggregate tables store precomputed results, which are measures that ha e been aggregated !typically summed# o er a set of dimensional attributes. "sing aggregate tables is a ery popular techni4ue for speeding up 4uery response times in decision support systems. o /f you are writing ).+ 4ueries or using a tool that only understands what physical tables e,ist !and not their meaning#, taking ad antage of aggregate tables and putting them to good use becomes more difficult as the number of aggregate tables increases. &he aggregate na igation capability of the )iebel Analytics )er er, howe er, allows 4ueries to use the information stored in aggregate tables automatically, without 4uery authors or 4uery tools ha ing to specify aggregate tables in their 4ueries. &he )iebel Analytics )er er allows you to concentrate on asking the right business 4uestion@ the ser er decides which tables pro ide the fastest answers. /Assume you are in B)) layer0 e !ave 1 dimension ta%les# in t!at# 2 ta%les need to !ave !ierarc!y# t!en in suc! a case is it mandatory to create !ierarc!ies for all t!e dimension ta%les? o -o, its not mandatory to define hierarchies to other Dimension tables. &an you !ave multiple data sources in 'ie%el Analytics? o Yes. How do you deal wit! case statement and e$pressions in sie%el analytics? o use e,pression builder to create case whenAthen.. end statement Do you know a%out Initiali3ation Blocks? &an you (ive me an e$ample w!ere you used t!em? o /nit blocks are used for instantiating a session when a user logs in. o &o create dynamic ariable you ha e to create /= to write s4l statement. w!at is .uery repository tool? o /t is utility of )eibel;$=/<< Admin tool o allows you to e,amine the repository metadata tool o for e,ample1 search for ob(ects based on name,type. o <,amine relationship between metadata ob(ects like which column in the presentation layer maps to which table in physical layer w!at is 4D5 and w!y do we need it? o Ba a De elopment Cit !BDC#, A software package that contains the minimal set of tools needed to write, compile, debug, and run Ba a applets. Oracle doesn6t recommend Opa.ue 7iews %ecause of performance considerations# so w!y,w!en do we use t!em? o an opa4ue iew is a physical layer table that consists of select statement. an opa4ue iew
should be used only if there is no other solution. &an you mi(rate t!e presentation layer to a different server. o -o we ha e to migrate the whole web D rpd files How do you identify w!at are t!e dimension ta%les and !ow do you decide t!em durin( t!e Business,Data modelin(? o Dimension tables contain descriptions that data analysts use as they 4uery the database. %or e,ample, the )tore table contains store names and addresses@ the :roduct table contains product packaging information@ and the :eriod table contains month, 4uarter, and year alues. < ery table contains a primary key that consists of one or more columns@ each row in a table is uni4uely identified by its primary-key alue or alues !y do we !ave multiple "+' in B)) layer? !at is t!e purpose? o to impro e the performance and 4uery response time. w!at is t!e full form of rpd? o there is no full form for rpd as such, it is (ust a repository file !8apidfile Database# !ow do i disa%le cac!e for only 2 particular ta%les? o in the physical layer, right click on the table there we will ha e the option which says cacheable How do you split a ta%le in t!e rpd (iven t!e condition? /+!e condition (iven was Broker and customer in t!e same ta%le0 'plit Broker and customer. o we need to make an alias table in the physical layer. !at type of protocol did you use in 'A'? o &C:;/: Did you create any new lo(ical column in B)) layer# !ow? o Yes. We can create new logical column in =MM layer. o <,ample1 8ight click on fact table -new logical column-gi e name for new logical column like &otal cost. o -ow in fact table source, we ha e one option column mapping, in that we can do all calculation for that new column. &an you use p!ysical join in B)) layer? o yes we can use physical (oin in =MM layer. when there is )CD type 5 we need comple, (oin in =MM layer. &an you use outer join in B)) layer? o yes we can. When we are doing comple, (oin in =MM layer ,there is one option type, outer (oin is there. !at are ot!er ways of improvin( summary .uery reports ot!er t!an A((re(ate 8avi(ation and &ac!e )ana(ement /nde,es Boin algorithm Mat; iew 4uery rewrite Web proper report design its optimal by making sure that it is not getting any addition column or rows
!at is level9%ase matrics? o +e al-base matrics means, ha ing a measure pinned at a certain le el of the dimension. %or <,ample, if you ha e a measure called Dollars, you can create a +e el =ased Measure called Yearly Dollars which !you guessed it# is Dollars for a Year. &his measure will always return the alue for the year e en if you drill down to a lower le el like 4uarter, monthA etc. &o create a le el based measure, create a new logical column based on the original measure !like Dollars in the e,ample abo e#. Drag and drop the new logical column to the appropriate le el in the Dimension hierarchy !in the abo e e,ample you will drag and drop it to Year in &ime Dim o A +=M is a metric that is defined for a specific le el or intersection of le els. o Monthly &otal )ales or .uarterly )ales are the e,amples. o You can compare monthly sales with 4uarterly sales. You can compare customer orders this 4uarter to orders this year