You are on page 1of 7

************************************************************************ INITIAL INTERVIEW QUESTIONS: 1. How many build methods are there?

There are 3 methods of builds, Using generation references, level references, parent child reference, adding list of new members. 2. Are you currently responsible for Production Essbase Cubes? Yes I am in production support and maintenance of Essbase cubes. a. How big is the largest cube? Ans: 8_____GB (If not GB, end interview) We have 2 native Essbase cubes and 6 planning cubes (AACorp_P, Apccd_P, AWA_P, NAPM_P, Europe_P, global_P) GL_Hist cube: 20 GB. AR_Credit has four native cubes: AR_NA : 15 GB AR_EUR : 13 GB AR_Global : 22 GB AR_AWA : 8 GB Planning application cubes: AACORP_P : 3 GB APCCD_P : 2 GB Packaging and consumer AWA_P : 2 GB Atlantic & Australi Global_P: 2GB Entire global NAPM_P: 3 GB North America Primary Metals Europe_P : 5 GB Europe Extrusions b. What is a datablock? Unique combination of Sparse standard dimension is a data block, which holds the atleast one value. The data block represents all the dense dimension members for its combination of sparse dimension members. Essbase creates an index entry for each data block. c. How do you order the outline? From the highest dense dimension to the lowest sparse dimension. And also by operators in outline, outline member formulas. d. How many processors on your server? HP-UX RP8400 16 processors portioned into 8 pr for 0P and 8 pr for 1p for performance issues. QA, Sand box Superdome with 54 processors added 16p/added 17 p UAT N4000 with 8 pr added 14 n 0d-RP8400-4 pr tnedd10n 1d-Superdome-training-tnedd08n Win2000-4 pr 16 pieces- prod planning, HAL, Reports, Analyzer, etc. Win2000-2 pr-7 pieces sand box planning, reports, analyzer etc. 3. Have you ever upgraded Essbase? Yes Involved in upgrade of Essbase from 6.0 to 6.5.4.0 Involved in upgrade of reports from 2.0 2.6 Involved in upgrade of Analyzer from 6.0 6.2.1 4. Name 3 Esscmd commands (excluding Login)? Some of ESSCMDs:

Create Cmds: CreateDB, CreateUser, CreateGroup, CreateApp, Load cmds: LoadApp, Loaddb, Loaddata, List cmds: Listdb, Listfiles, listFilters, ListApp, Delete cmds: DeleteUser, DeleteLog, DeleteApp, Deletedb Dba cmds: Export, Import, Update, Exit, Archive, beginArchive, EndArchive etc e. How many partition types are there: There are 3 types of partitions: 1. Replicated Partition 2. Transparent partition 3. Linked Partition 5. Why do you use Dynamic Calc? Dynamic calc to calculate the data value for a member as users requests it. Adv: Shortens regular database calculation times, reduces disk usage, reduces database restructure time. Disad: May increase retrieval time for the dynamically calculated data values. We have 2 types of dynamic calcs: Dynamic calc, Dynamic calc and Store 6. What members in the outline cannot be Dynamic Calc? Tagging Dense Dimension Members tag upper level members as dynamic calc, tagging level 0 members with simple formulas as dynamic calcs, do not members of dense dimensions as dynamic calc and store. Tagging Sparse Dimension Members upper level members of sparse dimension member, which have six or fewer children as dynamic calc or dynamic calc or store, complex formulas as dynamic calc or dynamic calc and store, do not tag upper level sparse members that have 20 or more descendants as dynamic calc. Tagging two pass members - Accounts dimensions (dense dimension) and Time dimension (dense dimension) members can be two pass members - to reduce amount of time needed to perform batch calculations can tag any dynamic calc or dynamic calc member as two pass even if the member is not on an accounts dimension. 7. For a member to be Two-Pass, what must exist in the outline? It should be an accounts dimension member in outline. Or use a calc script to perform calculation twice. 8. What functions are you familiar with: Name them? Essbase Functions: @ACCUM, @ABS, @ALIAS, @ANCESTORS, @AVG, @RANK, @CHILDREN, @COUNT, @Descendants, @Discount, @Interest, @ISMBR, @ISCHILD, @Max, @Min, @Prior, @PTD, @Rank, @Round, @Sum, @Sum, @Trend. f. Using the AVG function, if you want to leave out zeros, what parameter can you use? SKIPZERO, SKIPNONE, SKIPMISSING, SKIPBOTH

9. What has to exist in the outline before you can create an attribute dimension? Attribute can be text, Boolean, numeric and date. A standard dimension is any dimension that is not an attribute dimension. When an attribute dimension is associated with the standard dimension then the standard dimension is the base dimension to that attribute. Attribute dimension and members are dynamic cals Attribute dimension is not stored in database. All levels of the base dimension associated with the particular attribute dimension should be at the same level.

The level 0 members of the attribute dimensions are the only members that you can associate with the base dimension members. A base dimension can have many attributes but only one particular attribute from each attribute dimension.

10. How do you set up an attribute dimension in Essbase? List the steps? Add dimension and go through the properties and relate it to the sparse dimension 11. Misc. (Just ask a few) g. h. i. j. k. Have you used Hyperion Analyzer? y Have you used Hyperion Reports? y How many users use your application? From 50-1,000 Are you stronger at optimization or design? Optimization, well with design too. Do you like optimization or design better? Both

**********************************************************************

Essbase Interview Questions: 1. When Essbase server is crashed? How you are going to run it again? How you are going to check what made it crash? Srini answer: By using Start Essbase shell script or bat file on NT we will run it again. nohup Essbase b password. Which runs the Essbase in background. On Nt we run this Windows service. Ans: After a server crash, Essbase recovers the database, rolling back all transactions that were active when the interruption occurred. Recovery time depends upon the size of index. The larger the index, the longer it takes to recover the database. Essbase also recovers and consolidates free fragments (unused addressable units in data block). Essbase recovers data as soon as the server is started after a server interruption. Recovery consists of the following phases: o Transaction recovery rolls back all transactions that were active when the interruption occurred. o Index file recovery truncates files to the previously committed sizes. o Data free space recovery rebuilds the data free table spaces. The size of the index determines the duration of this phase. Only a media failure (faulty disk, disk failure etc) requires you to restore data from backups. Do not move copy/move/delete ess.ind, ess.pag, dbname.ind, dbname.esm or dbname.tct doing so result in data corruption. Dbname.tct database transaction control file which has entry for all transactions (Active, committed, aborted) carried on the database. Isolation Levels: Isolation determines how Essbase commits data to disk. When data is committed it is taken from server memory and written to database on disk. Essbase automatically commits the data to disk. There are 2 isolation levels for transactions: o Committed access committed access provides a high level of data consistency because only one transaction at a time can update data blocks. Under committed access Essbase retains redundant data until a transaction completes. Allow the

disk space for double the size of the database to accommodate redundant data. Here dead locks may occur. o Uncommitted access With uncommitted access (enabled by default) the Essbase kernel allows transactions to hold read/write locks on block-by-block basis. Essbase releases a block after it is updated but will not commit until a level is reached. You can make Essbase to perform explicit commit. Default 3000 but we had 10000. Commit blocks/ commit rows etc. With uncommitted access blocks are released more frequently than with committed access, when all blocks are locked until the end of transaction.

2. What is the difference between FIX and IF statements? In what situations they are used and what are its pros and cons? Srinians: FIX statement is to select the number blocks for particular set of calculations. IF is used to select the particular set of data cells with in a block to perform set of calculations. 3. What are different types of memory caches? What are their optimal memory settings? Srinians: Data cache, Index cache, Data file cache, Calc Cache, Dynamic calc cache, Retrieval Buffer cache 4. What are different servers used? What are its processor speeds, how many processors are there and memory details of each server? Srinians: HP UNIX 11i Superdome, RP 8400 8CPU 725 Mhz HP UNIX 11i Superdome 32 processors, 52 processors RP 8400 16 processors 32 GB memory N4000 8 processors 5. Cgf file settings. Where do you keep global variables for each application? What are different cfg settings used? JVMMODULELOACTION allows you to specify a specific java virtual machine library. AuthenticationModule specifies the elements needed for external authentication NO_HOSTNAME_LISTCONNECT: TRUE AGENTTHREADS : 15 specifies how many threads the agent may spawn-one thread is required for each initial connection to an application and database. SERVERTHREADS: 100 CalcCache : TRUE CalcCacheHigh: 50000000 CalcCachedefault: 5000000 CalcCacheLow:2000000 NetDelay:400 NetRetryCount:1200 SSLOGUNKNOWN: FALSE TIMINGMESSAGES: TRUE CALCLOCKBLOACKHIGH:500 CALCLOCKBLOCKDEFAULT:500 CALCLOCKBLOCKLOW:50 DLSINGLETHREADPERSTAGE: FALSE DLTHREADSPREPARE: 2 DLTHREADSWRITE:3 MULTIPLEBITMAPMEMCHECK: TRUE DATAERRORLIMIT: 5000

QUICKLOGIN: ON MAXLOGINS: 15000 Database settings, application setting details. 6. Some MAXL commands, ESSCMD Commands. MAXL Commands: Create group: Create application, Create calculation, create database, create filter, create function, create group, Create macro, create users etc. Alter group: alter application, alter database, alter group, alter user, and alter system. Display group: display application, display database, display filter, display group, display system, display session, display user etc. Drop: drop application, drop calculation, drop database, drop filter, drop function, drop group, drop user. Other group: export data, import data, execute calculation, import dimension etc. ESSCMD Commands: Create group: Createapp, createdb, creategroup, createuser, createlocation. Delete group: deleteapp, deletedb, deletegroup, deletelog, deleteuser, disablelogin, enablelogin. Get group:getappinfo, getappstate, getdbinfo, getdbstate, getmembers,getdbstats etc List group: listapp, listdb, listgroup, listlogins, listusers. Export, import, sleep, shutdownserver, unloadapp, unloaddb etc. 7. When I/O is overwhelmed? What can be done? Why it is overwhelmed? Srinians: Use Buffered IO instead of Direct IO and increase the data and index cache setting to reduce the IO. 8. What are different ways of calcscript optimization? Srini can you pl answer this 9. What are different database settings? Why are their advantages? 10. What are different types of build? What are different types of data loads? 11. What is the limitation of generation build? Srini can you pl answer this 12. What is a datablock? Inside the datablock how the data cells were arranged? Srinians: Data block is intersection of existing sparse member intersection, which is identified by index number. With in the each block the each dense member intersection is stored as cell, which holds actual data numbers. The cells dont have any index and Essbase engine searches for these cells with internal search mechanism. 13. When a planning server is down? What are the sequences of steps used to bring it to running? i.e., how different servers are started in sequence? Srinians: Start Planning web server service, HBR service and kill any processes started with Hsx* 14. When users are accessing an application, dba wants to run a batch job with out affecting the users? How this can be done. Srinians: Depends, Batch can be run on a working cube and swap it after the process like AR or else do the sparse calculations which dont impact users. 15. Length of batch job scripts? How big are there may be in no of lines etc. There is no limit on the no of lines in a batch script.

========================================== Vish Interview questions & Answers: 16. Differences between stored procedures and packages. 17. Difference between data block and data block size. Data block is a block which essbase creates for every unique combination of sparse dimensions. Each data block represents data from the dense dimensions. Data blocks are likely to have few empty cells. Each data block represents all dense dimensions member values for that unique combination of sparse dimension members. A data block size of 8KB to 100 KB provides optimal performance in most cases. 18. Difference between ESScmds and Maxl commands. 19. How do you invoke esscmd? By typing esscmd at the command prompt. 20. 21. 22. 23. 24. What do you mean by gigantic cube? How do you tune them? Difference between essbase 6.5 & essbase 7.1 Hyperion Analyzer Hyperion EIS Have you worked on production server Yes.

25. How do you see the users that are logged on? From connections Menu. In 71 we see it in sessions. 26. Some calculation commands. Some of Calculation Commands: Calcall, Clearblock, Cleardata, Agg, datacopy, Fix & EndFix, If and Endif, Set commandsSet Calclockblock, Set cache, Set notice etc 27. Where is calcscript file saved, server or client, database entry or file. It could be saved in server machine and also in client machine. 28. How do you call calcscript from Analyzer. 29. How do you connect to analyzer to the database? By creating a database connection and will map it to the report. 30. What is a data block? Data block is a block, which has data values for all dense dimensions for those unique sparse dimension combinations. 31. How did you get data? By text files from legacy systems, from oracle and excel spreadsheets too. 32. How do you calculate sparse members in an outline? Placing the sparse dimension, which has fewer members first and then placing the next highest etc. Order of calculation: Dimension tagged as Accounts Dimension tagged as time Dense dimension Sparse dimension

Optimize data loads Grouping sparse member combinations Positioning data same as outline Making data source as small as possible Making source fields as small as possible Loading data from server itself.

You might also like