Best Practices for SAP BI using DB2 9 for z/OS - IBM Redbooks

More documents

Recommendations

Info

processors in terms of macros that consist of several method calls manipulating the structures of queries. The BI Java SDK provides two command processors: ► OLAP Command Processor, for manipulating OLAP queries ► Relational Command Processor, for manipulating relational queries Common Client Interface An API defined by Sun's JCA specification that is common across heterogeneous EISs. It is designed to be “tool-able”—that is, it leverages the Java Beans architecture so that development tools can incorporate the CCI into their architecture. Note that the BI Java Connectors implement only the connection interfaces defined by the CCI. The CCI's interaction interfaces, data interfaces, and metadata interfaces, however, are not implemented by the BI Java SDK. BI-specific client APIs that are tailored for OLAP interactions are provided by the BI Java Connectors. common warehouse metamodel A standard that is recognized by the Object Management Group (OMG) and that describes the exchange of metadata in the following areas: ► Data Warehousing ► Business Intelligence ► Knowledge Management ► Portal Technology CWM uses ► UML to model metadata ► MOF to access metadata ► XMI to exchange metadata You can find the current CWM specification on the OMG home page. communication structure The communication structure is independent of the source system and has the structure of an InfoSource. It contains all the InfoObjects that belong to an InfoSource. All the data is updated into the InfoCube with this structure. The system always refers to the actively saved version of the communication structure. Technically (that is, with respect to length, type) the fields of the communication structure correspond to the InfoObjects of the Business Information Warehouse. 292 Best Practices for SAP BI using DB2 9 for z/OS
comparison column A write-protected data column in the planning layout. These comparison columns serve to display data for specific characteristic value combinations for which no data is to be entered in the planning layout. Comparison columns can be used as a reference for the plan data that is being entered (for example, when planning for the next quarter, the comparison column could display the actual data from the previous quarter). For planning layouts that contain key figures in the lead column, what is said about columns also applies to rows (comparison rows). condenser A program that compresses the contents of an InfoCubE- fact table. Connector Gateway An SAP Enterprise Portal service that provides instances of connections to Portal components. control query A help query that you execute before you execute queries in the Web template. You use the result of this query to parameterize the Web template. Cube A set of data organized as a multidimensional structure defined according to dimensions and measures. Related SAP BI terms include InfoCube and query. Crystal Enterprise Server component that is required to publish formatted reports created with the Crystal Reports Designer on the basis of BI data. Formatted reports that are stored on the Crystal Enterprise server can be called up from there and displayed on the Web. Content and user administration is carried out as part of the integration using the BI server. Crystal Report BI object type. Report definition for a formatted report that has been created using the Crystal Reports Designer on the basis of a BI query. The report definition is saved in SAP BI and published to the Crystal Enterprise server. Appendix A. Glossary for SAP BI 293
Page 1:
Best Practices for SAP BI using DB2
Page 4 and 5:
Note: Before using this information
Page 6 and 7:
4.4 Extended sparse index and in-me
Page 8 and 9:
11.3 SAP BI database and InfoCube l
Page 10 and 11:
viii Best Practices for SAP BI usin
Page 12 and 13:
Trademarks The following terms are
Page 14 and 15:
The team that wrote this IBM Redboo
Page 16 and 17:
Thanks to these authors from the Pr
Page 18 and 19:
► Send your comments in an e-mail
Page 20 and 21:
1.1 Product overview The business w
Page 22 and 23:
In general, database administrators
Page 24 and 25:
► Horizontal growth is achieved b
Page 26 and 27:
8 Best Practices for SAP BI using D
Page 28 and 29:
2.1 BI architecture Figure 2-1 show
Page 30 and 31:
New term Old term Analysis authoriz
Page 32 and 33:
negative values means missing stati
Page 34 and 35:
- When calculating database statist
Page 36 and 37:
If you specify a partitioning schem
Page 38 and 39:
Important: Never partition a PSA ta
Page 40 and 41:
Figure 2-10 DPSIs versus NPSIs Figu
Page 42 and 43:
Finally, Figure 2-12 explains which
Page 44 and 45:
Figure 2-14 shows how the snowflake
Page 46 and 47:
► Realignment This refers to the
Page 48 and 49:
3.1 Base recommendations - SAP BI i
Page 50 and 51:
the F- fact table, the E- fact tabl
Page 52 and 53:
34 Best Practices for SAP BI using
Page 54 and 55:
4.1 Dynamic index ANDing for star s
Page 56 and 57:
Figure 4-2 and Figure 4-3 on page 3
Page 58 and 59:
4.2.1 Processing RUNSTATS for BI ta
Page 60 and 61:
RUNSTATS will then be called over t
Page 62 and 63:
DSNU610I -DB2T 081 16:36:03.85 DSNU
Page 64 and 65:
Often the single-value based freque
Page 66 and 67:
Several things worth noting are:
Page 68 and 69:
The unit of a value specified to MX
Page 70 and 71:
condition, generated by SAP BI upon
Page 72 and 73:
There is a separate IBM Redbooks pu
Page 74 and 75:
5.1 Load PSA In this section we dis
Page 76 and 77:
5.1.1 Load PSA To load data into th
Page 78 and 79:
After the source system, the DataSo
Page 80 and 81:
Note the available tabs (registers)
Page 82 and 83:
Figure 5-8 shows the provided optio
Page 84 and 85:
Figure 5-10 shows the following:
Page 86 and 87:
The PSA table name is created autom
Page 88 and 89:
5.1.4 PSA partitioning Partitioning
Page 90 and 91:
Page 92 and 93:
An InfoCube, whether data target or
Page 94 and 95:
6.1 Types of InfoCubes Data targets
Page 96 and 97:
After selecting at least one time d
Page 98 and 99:
6.2.1 Repartitioning implementation
Page 100 and 101:
You will then be presented with a w
Page 102 and 103:
From within the monitor you are abl
Page 104 and 105:
Once your repartitioning process is
Page 106 and 107:
Figure 6-14 is displayed once you h
Page 108 and 109:
When you remodel an InfoProvider yo
Page 110 and 111:
Figure 6-17 demonstrates how you wo
Page 112 and 113:
At this step of the process you are
Page 114 and 115:
6.3.2 Benefits of remodeling InfoPr
Page 116 and 117:
6.4.3 Cumulative data versus non-cu
Page 118 and 119:
Important: After the initial compre
Page 120 and 121:
► Use DPSIs for the cube’s E- f
Page 122 and 123:
7.1 Differences between DataStore a
Page 124 and 125:
Figure 7-1 shows the initial status
Page 126 and 127:
4. By selecting the Create button,
Page 128 and 129:
Partitioning and clustering are bot
Page 130 and 131:
7.4.2 Choosing and changing the clu
Page 132 and 133:
7.4.4 Partitioning a DataStore obje
Page 134 and 135:
From the ABAP Dictionary: Utility f
Page 136 and 137:
Figure 7-9 DB2/390 Storage Attribut
Page 138 and 139:
partitioning was used. The CREATE T
Page 140 and 141:
, "/BIC/RECORDMODE" POSITION( 00289
Page 142 and 143:
Upon completion of the job, you wil
Page 144 and 145:
| "/BIC/DIVISION" VARGRAPHIC (00000
Page 146 and 147:
| sql: | DELETE FROM DDSTORAGE | WH
Page 148 and 149:
| "/BIC/CRMEM_VAL" DECIMAL (000017,
Page 150 and 151:
| DDL time(___8): .........4 millis
Page 152 and 153:
Page 154 and 155:
8.1 Troubleshooting general system
Page 156 and 157:
You need to ascertain: ► Which us
Page 158 and 159:
Further information about the subje
Page 160 and 161:
8.2 Analyzing load performance In t
Page 162 and 163:
We use transaction ST03 → Expert
Page 164 and 165:
Click the Details tab of the RSMO w
Page 166 and 167:
Select the component of interest to
Page 168 and 169:
Figure 8-11 All statistical records
Page 170 and 171:
Figure 8-13 Single statistical reco
Page 172 and 173:
Loading from a flat file You can av
Page 174 and 175:
and right-click to select the Simul
Page 176 and 177:
Note: You must activate the saving
Page 178 and 179:
5. In the right-hand panel of the w
Page 180 and 181:
If your query has long run times th
Page 182 and 183:
that the query runs against → Mai
Page 184 and 185:
If the main component of query perf
Page 186 and 187:
► Display Statistics Data (to sho
Page 188 and 189:
Select the appropriate button, ente
Page 190 and 191:
Figure 8-29 Statement details To ch
Page 192 and 193:
The ratio (ST04 statement cache “
Page 194 and 195:
Figure 8-32 Visual Explain - SQL te
Page 196 and 197:
8.5 SAP Query Analysis Tool The SAP
Page 198 and 199:
Display of sql access path SQL quer
Page 200 and 201:
The access path is displayed hierar
Page 202 and 203:
The table information tab shown in
Page 204 and 205:
DB2 Plan Table Figure 8-42 DB2 plan
Page 206 and 207:
Graphical view of each query block
Page 208 and 209:
Now you see the graphical view of t
Page 210 and 211:
8.5.4 Example of using the query an
Page 212 and 213:
Page 214 and 215:
9.1 The BIA architecture Figure 9-1
Page 216 and 217:
Figure 9-2 BIA t-shirt sizing The B
Page 218 and 219:
From the hardware side, the BIA is
Page 220 and 221:
9.7 Overview The improvement of the
Page 222 and 223:
10.1 SQL efficiency for DSO Queries
Page 224 and 225:
The first join predicate is "/BIC/A
Page 226 and 227:
Figure 10-5 Associate local predica
Page 228 and 229:
Step 3 is repeated for each master
Page 230 and 231:
Using the InfoCube query example in
Page 232 and 233:
The first join predicate in Figure
Page 234 and 235:
Thus, we have broken down the origi
Page 236 and 237:
Figure 10-14 Comparing filtering Th
Page 238 and 239:
AND "/BIC/AZOSASALE00"."PLANT" = "/
Page 240 and 241:
Column Local or join CALMONTH Local
Page 242 and 243:
10.4 Indexing of InfoCubes The defa
Page 244 and 245:
Page 246 and 247:
11.1 Background To be more competit
Page 248 and 249:
11.2.1 Hardware and software config
Page 250 and 251:
11.2.2 Network infrastructure IBM S
Page 252 and 253:
the chassis for blade-to-blade conn
Page 254 and 255:
As part of the pre-benchmark planni
Page 256 and 257:
11.4 Test case scenarios The goal o
Page 258 and 259:
Figure 11-3 explains the data flow
Page 260 and 261: In addition, each cube, especially
Page 262 and 263: Measurement results Table 11-2 prov
Page 264 and 265: Figure 11-7 shows the average CPU u
Page 266 and 267: The number of queries per hour show
Page 268 and 269: Figure 11-10 shows that the average
Page 270 and 271: 11.5.1 Recommendations for configur
Page 272 and 273: 254 Best Practices for SAP BI using
Page 274 and 275: 12.1 Terminology: BI objects and pr
Page 276 and 277: Metadata plays a fundamental role i
Page 278 and 279: 12.3 Dataflow in SAP BI Data flow i
Page 280 and 281: 12.4 Information access As previous
Page 282 and 283: Predecessor Successor WW Sales Euro
Page 284 and 285: An InfoCube in SAP BI can have up t
Page 286 and 287: 12.8 Aggregates 4. The dimension ta
Page 288 and 289: The next example is an aggregate th
Page 290 and 291: For example, assume that different
Page 292 and 293: can be updated into InfoCubes or ot
Page 294 and 295: Figure 12-11 shows the steps that o
Page 296 and 297: Figure 12-12 on page 279 shows the
Page 298 and 299: The content, depicted in Figure 12-
Page 300 and 301: considerably more data records are
Page 302 and 303: The nodes stand for data sources, t
Page 304 and 305: values, hierarchies, hierarchy node
Page 306 and 307: BI SAP Query Connector A resource a
Page 308 and 309: Business Intelligence Cockpit Web-b
Page 312 and 313: Crystal Reports Designer Design com
Page 314 and 315: Similar to the OLAP-standard term c
Page 316 and 317: P table Master data table for time-
Page 318 and 319: The process instance contains the m
Page 320 and 321: Relational Query Model An abstracti
Page 322 and 323: SAP Business Information Warehouse
Page 324 and 325: source object element Component of
Page 326 and 327: elements to be displayed in the Web
Page 328 and 329: Web application object catalog Web
Page 330 and 331: You insert one or more queries in t
Page 332 and 333: 314 Best Practices for SAP BI using
Page 334 and 335: IT scenario The environment we used
Page 336 and 337: Checks under z/OS and DB2 We recomm
Page 338 and 339: To check that the SAPCL procedure w
Page 340 and 341: List of SAP BW OSS Notes Table C-1
Page 342 and 343: SAP BI on DB2 for z/OS 390016 BW Da
Page 344 and 345: Network configuration: static multi
Page 346 and 347: START CU091060 START CU0A1070 START
Page 348 and 349: Table name Average row length Total
Page 350 and 351: Table name Average row length Total
Page 352 and 353: Parameters BPK16K0 BPK16K1 VIRTUAL
Page 354 and 355: drwxr-xr-x 3 t11adm sapsys 2048 200
Page 356 and 357: Example D-6 shows a cube index crea
Page 358 and 359: B49/S_RTNSVAL 24 DECIMAL 17 2 N 4 N
Page 360 and 361:
-rw-r--r-- 1 t11adm sapsys 2084482
Page 362 and 363:
-rw-r--r-- 1 t11adm sapsys 547449 2
Page 364 and 365:
Page 366 and 367:
Online resources These Web sites ar
Page 368 and 369:
Help from SAP SAP Support and downl
Page 370 and 371:
C Candidates for partiioning Data W
Page 372 and 373:
M master data 155, 211, 258, 277 si
Page 374 and 375:
star schema 265-266 also see InfoCu
Page 376:
Best Practices for SAP BI using DB2
show all

Best Practices for SAP BI using DB2 9 for z/OS - IBM Redbooks

Create successful ePaper yourself

Delete template?

Save as template?