Download - DW Implementation Boot Camp - Student Manual
-
5/22/2018 DW Implementation Boot Camp - Student Manual
1/118
For Oracle employees and authorized partners only. Do not distribute to third parties. 2008 Oracle Corporation Proprietary and Confidential
For Oracle employees and authorized partners only. Do not distribute to third parties. 2009 Oracle Corporation Proprietary and Confidential
The following is intended to outline our generalproduct direction. It is intended for informationpurposes only, and may not be incorporated into anycontract. It is not a commitment to deliver anymaterial, code, or functionality, and should not berelied upon in making purchasing decisions.The development, release, and timing of anyfeatures or functionality described for Oraclesproducts remains at the sole discretion of Oracle.
Safe Harbor Statement
For Oracle employees and authorized partners only. Do not distribute to third parties. 2009 Oracle Corporation Proprietary and Confidential
Use of this Site (Site) or Materials constitutes agreement with the following terms and conditions:
1. Oracle Corporation (Oracle) is pleased to allow its business partner (Partner) to download andcopy the information, documents, and the online training courses (collectively, Materials") found on thisSite. The use of the Materials is restricted to the non-commercial, internal training of the Partnersemployees only. The Materials may not be used for training, promotion, or sales to customers or otherpartners or third parties.
2. All the Materials are trademarks of Oracle and are proprietary information of Oracle. Partner or otherthird party at no time has any right to resell, redistribute or create derivative works from the Materials.
3. Oracle disclaims any warranties or representations as to the accuracy or completeness of anyMaterials. Materials are provided "as is" without warranty of any kind, either express or implied,including without limitation warranties of merchantability, fitness for a particular purpose, and non-infringement.
4. Under no circumstances shall Oracle or the Oracle Authorized Delivery Partner be liable for any loss,damage, liability or expense incurred or suffered which is claimed to have resulted from use of this Siteof Materials. As a condition of use of the Materials, Partner agrees to indemnify Oracle from and againstany and all actions, claims, losses, damages, liabilities and expenses (including reasonable attorneys'fees) arising out of Partners use of the Materials.
5. Reference materials including but not limited to those identified in the Boot Camp manifest can not beredistributed in any format without Oracle written consent.
Oracle Training Materials Usage
Agreement
For Oracle employees and authorized partners only. Do not distribute to third parties. 2008 Oracle Corporation Proprietary and Confidential
Title of Presentation
Presenters NamePresenters Title
Data Warehouse course introduction
-
5/22/2018 DW Implementation Boot Camp - Student Manual
2/118
For Oracle employees and authorized partners only. Do not distribute to third parties. 2009 Oracle Corporation Proprietary and Confidential
Module Agenda
Introductions
Introduction to data warehousing
Creating business value from IT
Data warehouse architecture & design
Preparing for the hands-on labs
For Oracle employees and authorized partners only. Do not distribute to third parties. 2009 Oracle Corporation Proprietary and Confidential
Introductions
The goals of this course
Provide background to understand technicalchallenges presented by data warehouses
Provide a technical overview of the Oracle Database11g Data Warehousing features with extensivehands-on labs
Guidelines on sizing of Data Warehouse hardwareconfigurations
Share best practices for deploying Oracle DataWarehouse architectures and platforms
For Oracle employees and authorized partners only. Do not distribute to third parties. 2009 Oracle Corporation Proprietary and Confidential
Introductions
Who are you? Name
Company
Role
What is your prior experience? Database administration
Data warehouse experience
10g or 11g Database experience
How do you expect to benefit from this course?
For Oracle employees and authorized partners only. Do not distribute to third parties. 2009 Oracle Corporation Proprietary and Confidential
-
5/22/2018 DW Implementation Boot Camp - Student Manual
3/118
Introductions
Who am I?
For Oracle employees and authorized partners only. Do not distribute to third parties. 2009 Oracle Corporation Proprietary and Confidential
Course modules
Introduction to data warehousing
Introduction to Oracle data warehouse features
Parallel execution - Lab
Partitioning - Lab
Result Cache - Lab
OLAP - Lab
Advanced compression - Lab
Data integration and ETL- Demo
Course modulesContinued
Sizing Oracle data warehouse platforms
Exadata overview
Best practices for performance
For Oracle employees and authorized partners only. Do not distribute to third parties. 2009 Oracle Corporation Proprietary and Confidential
Introduction to data
warehouses
-
5/22/2018 DW Implementation Boot Camp - Student Manual
4/118
For Oracle employees and authorized partners only. Do not distribute to third parties. 2009 Oracle Corporation Proprietary and Confidential
All about data warehousesWhat is a data warehouse?
A data warehouse is A repository of data
From multiple sources
Used for analysis
Strategic value is realized from data aggregation
For Oracle employees and authorized partners only. Do not distribute to third parties. 2009 Oracle Corporation Proprietary and Confidential
All about data warehousesWhy have a data warehouse?
Increase the value inherent in data
Extract strategic information from tactical data
Drive IT value into business results
For Oracle employees and authorized partners only. Do not distribute to third parties. 2009 Oracle Corporation Proprietary and Confidential
All about data warehousesHow is a data warehouse different from OLTP?
Read-mainly
Typically, real-time not required
Large data sets
Complex queries based on many factors
For Oracle employees and authorized partners only. Do not distribute to third parties. 2009 Oracle Corporation Proprietary and Confidential
All about data warehousesHow is a data warehouse the same as OLTP?
Modern data warehouses Near real time
Mixed workloads
ETL
OLTP
Write operations make OLTP issues relevant
Performance
Locking
-
5/22/2018 DW Implementation Boot Camp - Student Manual
5/118
For Oracle employees and authorized partners only. Do not distribute to third parties. 2009 Oracle Corporation Proprietary and Confidential
All about data warehousesUsing a data warehouse
Data
WarehouseCustomers Web
Servers
Personalization
Content
Servers
Fulfillment
Centers
Supply
Chain
Planning
Apps
Financial
Analysis
Campaign
Generation Email
Servers
Internal
End Users
Partners
For Oracle employees and authorized partners only. Do not distribute to third parties. 2009 Oracle Corporation Proprietary and Confidential
All about data warehousesUsing a data warehouse Amazon.com
0
100
200
300
400
500
600
700
Terabytes
2000
2001
2002
2003
2004
2005
2006
2007
2008
2009
2010
NA
FE
EU
Data Volume has grown 2x year over year for the past 4 years.Projecting higher growth in upcoming years*:
- additional lines of business / product lines supported- huge standard reporting growth with more partners supported
For Oracle employees and authorized partners only. Do not distribute to third parties. 2009 Oracle Corporation Proprietary and Confidential
Creating business value
from IT
Creating business value from IT
IT PRIORITIES BUSINESS PRIORITIES
Reduced Cost Business Value
Improved Security Ease of Access
Improved SystemPerformance
Results in Timefor Deadlines
Manageability /Consistency
Good Enough Infrastructure
For Oracle employees and authorized partners only. Do not distribute to third parties. 2009 Oracle Corporation Proprietary and Confidential
-
5/22/2018 DW Implementation Boot Camp - Student Manual
6/118
Creating business value from IT
Increasing Return on Investment for informationproducers by Reducing current costs of analysis for lines of business
Empower LOBs with tools, applications, and automatedanalysis rather than manually creating custom reports
Reducing ongoing costs of analysis for IT
Faster and simpler development of analytic applications
Centralized management
For Oracle employees and authorized partners only. Do not distribute to third parties. 2009 Oracle Corporation Proprietary and Confidential
Creating business value from IT
Increasing Return on Investment for informationconsumers in lines of business by Enabling top-line growth
New customers/markets, best customers,
React faster than the competition
Enabling bottom-line savings
Optimize operations, mitigate risk,
Better manage the business than the competition
For Oracle employees and authorized partners only. Do not distribute to third parties. 2009 Oracle Corporation Proprietary and Confidential
For Oracle employees and authorized partners only. Do not distribute to third parties. 2009 Oracle Corporation Proprietary and Confidential
Data warehouse
architecture and design
For Oracle employees and authorized partners only. Do not distribute to third parties. 2009 Oracle Corporation Proprietary and Confidential
Data warehouse architecture and designArchitecture
Data warehouse
Operational data store
Data marts
Federated data marts
-
5/22/2018 DW Implementation Boot Camp - Student Manual
7/118
For Oracle employees and authorized partners only. Do not distribute to third parties. 2009 Oracle Corporation Proprietary and Confidential
Data warehouse architecture and designData design
Analysis performed on a set of facts Examined through different dimensions
Usually has time dimension
Queries typically examine facts with many dimensions
For Oracle employees and authorized partners only. Do not distribute to third parties. 2009 Oracle Corporation Proprietary and Confidential
Data warehouse architecture and designDesign options
Star schema
Snowflake schema
OLAP cubes
Hybrid organization
For Oracle employees and authorized partners only. Do not distribute to third parties. 2009 Oracle Corporation Proprietary and Confidential
Data warehouse architecture and designDesign options Star schema
SalesSales
DistributorDistributorOEMOEM
ChannelChannel Geography Geography
TypeType
SizeSize
ColorColor
ProductProduct TimeTime
YearYear
QuarterQuarter
MonthMonth
RegionRegion
DistrictDistrictBranchBranch
Fact Table:Fact Table:
SalesSales
TransactionsTransactions
For Oracle employees and authorized partners only. Do not distribute to third parties. 2009 Oracle Corporation Proprietary and Confidential
Data warehouse architecture and designDesign options Snowflake schema
-
5/22/2018 DW Implementation Boot Camp - Student Manual
8/118
For Oracle employees and authorized partners only. Do not distribute to third parties. 2009 Oracle Corporation Proprietary and Confidential
Data warehouse architecture and designDesign options OLAP cubes
Product
Sales Cube
Inventory Cube
SALES FACTSALES FACT
TIME
INVENTORYINVENTORY
REGION
CUST
ITEM
TimeGeography
Product
Time
For Oracle employees and authorized partners only. Do not distribute to third parties. 2009 Oracle Corporation Proprietary and Confidential
Data warehouse architecture and designDesign options Hybrid organization
Combination of structures Star schema
OLAP cubes
Third normal form
Oracle query optimizer can handle appropriately forall structures
For Oracle employees and authorized partners only. Do not distribute to third parties. 2009 Oracle Corporation Proprietary and Confidential
Lab Setup
Lab setup - Connection
Instructions for dialing in to hosted image
For Oracle employees and authorized partners only. Do not distribute to third parties. 2009 Oracle Corporation Proprietary and Confidential
-
5/22/2018 DW Implementation Boot Camp - Student Manual
9/118
Lab setup - Testing
Login to the VM Linux environment with oracle/oracle
Open a new command window
Start a SQL Plus session as SH/SH
For Oracle employees and authorized partners only. Do not distribute to third parties. 2009 Oracle Corporation Proprietary and Confidential
Lab setup VM environment overview
All passwords for OS & Oracle logins are same as theusername
Except - root, SYS, SYSTEM, and SYSMAN areoracle
All the lab scripts are stored in the/home/oracle/wkdirdirectory
Few shortcuts:
Command
Window
Firefox
Browser
Analytic
Workspace
Manager
SQL
Developer
For Oracle employees and authorized partners only. Do not distribute to third parties. 2009 Oracle Corporation Proprietary and Confidential
Summary
A data warehouse is a way to provideadditional value from existing data
Data warehouse architectures can includeoperational data stores, federated datamarts and data marts
Data warehouses can use star schemas,snowflake schema, OLAP cubes, thirdnormal form or a mixture of differentstructures
For Oracle employees and authorized partners only. Do not distribute to third parties. 2009 Oracle Corporation Proprietary and Confidential
For Oracle employees and authorized partners only. Do not distribute to third parties. 2008 Oracle Corporation Proprietary and Confidential
-
5/22/2018 DW Implementation Boot Camp - Student Manual
10/118
For Oracle employees and authorized partners only. Do not distribute to third parties.
2008 Oracle Corporation Proprietary and ConfidentialFor Oracle employees and authorized partners only. Do not distribute to third parties.
2008 Oracle Corporation Proprietary and Confidential
For Oracle employees and authorized partners only. Do not distribute to third parties. 2008 Oracle Corporation Proprietary and Confidential
Title of Presentation
Presenters NamePresenters Title
Data Warehouse and Business Intelligenceoverview
For Oracle employees and authorized partners only. Do not distribute to third parties. 2009 Oracle Corporation Proprietary and Confidential
Module Agenda
Why Oracle for data warehousing?
Core data warehousing features
Oracle 11g data warehousing features
Oracle hardware data warehouse solutions
Analytics and data mining
Oracle Business Intelligence Platform (OBIEE)
-
5/22/2018 DW Implementation Boot Camp - Student Manual
11/118
For Oracle employees and authorized partners only. Do not distribute to third parties. 2009 Oracle Corporation Proprietary and Confidential
Why Oracle for data
warehousing?
Why Oracle for data warehousing?
Best-of-breed for data warehouses and data marts
Consolidation platform Eliminates the costs and inefficiencies of multiple data stores
Complete, integrated ELT and data-qualitycapabilities
Unmatched analytic capabilities Embedded OLAP, data-mining and statistics all accessible via
SQL
Optimized for grid computing
Next-generation BI with integrated OLTP and DSS
Optimized hardware solutions
Foundation of Oracle Business Intelligence solutions
For Oracle employees and authorized partners only. Do not distribute to third parties. 2009 Oracle Corporation Proprietary and Confidential
Oracle - #1 in data warehouse market
share
For Oracle employees and authorized partners only. Do not distribute to third parties. 2009 Oracle Corporation Proprietary and Confidential
Core data warehousing
features
-
5/22/2018 DW Implementation Boot Camp - Student Manual
12/118
Oracle data warehouse platform
DW Services
Administration
Workload Mgmt
Self-Tuning
Security
Virtual PrivateDatabase
Database Vault
Data Integration
Bulk ETL + Real-TimeETL
Data Quality
Metadata ManagementExtensibleRepository
Lineage and ImpactAnalysis
Brainy Software
Analytics
Multi-DimensionalCalculations
Statistics Data Mining
Intelligent Query Processing
MaterializedViews
BitmapIndexes
PartitionElimination
Star QueryOptimization
Brawny Hardware
Scalable Data Management
ParallelExecution
Partitioning RACAutomatic
StorageMgmt
Compression
Scalable Hardware Infrastructure
ReferenceConfigurations
OptimizedWarehouse
ExadataDatabaseMachine
ExadataStorage Server
For Oracle employees and authorized partners only. Do not distribute to third parties. 2009 Oracle Corporation Proprietary and Confidential
Core data warehouse featuresBit-mapped indexes
PARTStable 001001 GREENGREEN MEDMED 98.198.1
002002 REDRED MEDMED 12411241003003 REDRED SMALLSMALL 100.1100.1004004 BLUEBLUE LARGELARGE 54.954.9005005 REDRED MEDMED 124.1124.1006006 GREENGREEN SMALLSMALL 60.160.1...... ........ .......... ......
sizesize == SMALLSMALL 0 0 1 0 1 1 0 1 0 1 0 0 0 1 0 10 0 1 0 1 1 0 1 0 1 0 0 0 1 0 1sizesize == MEDMED 1 0 0 0 0 0 1 0 0 0 0 1 0 1 0 01 0 0 0 0 0 1 0 0 0 0 1 0 1 0 0sizesize == LARGELARGE 0 1 0 1 0 0 0 0 1 0 1 0 1 0 1 00 1 0 1 0 0 0 0 1 0 1 0 1 0 1 0
Index on COLOR
colorcolor == BLUEBLUE 0 1 0 1 0 0 1 0 1 0 1 0 0 0 1 00 1 0 1 0 0 1 0 1 0 1 0 0 0 1 0colorcolor == REDRED 0 0 1 0 0 0 0 1 0 0 0 0 1 0 0 10 0 1 0 0 0 0 1 0 0 0 0 1 0 0 1colorcolor == GREENGREEN 1 0 0 0 1 1 0 0 0 1 0 1 0 1 0 01 0 0 0 1 1 0 0 0 1 0 1 0 1 0 0
010
010
010
010
partnopartno color color sizesize weightweight
Index on SIZE
SELECT count(*)
FROM parts
WHERE
size = MED a AND
color = RED
For Oracle employees and authorized partners only. Do not distribute to third parties. 2009 Oracle Corporation Proprietary and Confidential
Core data warehouse featuresParallel bit-mapped star joins
Dramatic performance gains
Innovative use of bitmap indexes
Ideal for complex star schemas Multiple fact tables
Many dimensions
Unconstrained dimensions
Large dimension tables
Optimized for sparse fact tables
Parallel execution
Optimizer modifies an SQL query for fasterperformance with star schema
Bitmapped indexes used to select rows from facttables Replaces multiple joins
Results are joined to dimension tables for additionalinformation
Must enable in database and conform to constraints
Core data warehouse featuresStar transformation
-
5/22/2018 DW Implementation Boot Camp - Student Manual
13/118
For Oracle employees and authorized partners only. Do not distribute to third parties. 2009 Oracle Corporation Proprietary and Confidential
Core data warehouse featuresMaterialized views
What were salesWhat were salesin the West andin the West and
South regions forSouth regions for
the last threethe last three
quarters?quarters?
x
(2 GB)
SALESSALESSALESSALES
SALESSALESPERPER
MONTHMONTHBY REGIONBY REGION
SALESSALESPERPER
MONTHMONTHBY REGIONBY REGION
Stored summaries aremaintained for fasterwarehouse queryprocessing
For Oracle employees and authorized partners only. Do not distribute to third parties. 2009 Oracle Corporation Proprietary and Confidential
Core data warehouse featuresMaterialized views query rewrite
SELECT p.brand, r.country, t.month,SUM(s.amt) tot_sales
FROM sales s, region r,time t, product p
WHERE s.region_id = r.region_idAND s.sdate = t.curdateAND s.prod_code = p.prod_codeGROUP BY p.brand, r.country, t.monthHAVING SUM(s.amt) > 5000000;
SELECT brand, country, month, tot_salesFROM sales_sumryWHERE tot_sales > 5000000;
Optimizer
transparently
rewrites queryon detailed
data ...
Optimizer
transparently
rewrites queryon detailed
data ...
to access
data in the
materialized
view
to access
data in the
materialized
view
For Oracle employees and authorized partners only. Do not distribute to third parties. 2009 Oracle Corporation Proprietary and Confidential
Core data warehouse featuresMaterialized views data refresh
Maintain the timeliness of storedresult to preference
Several options for refreshing datain materialized views
Manage the effect of performanceon operational data
No impact on load performance
SUMSUMORDERSORDERSbyby
QUARTERQUARTER
COUNTORDERS
byPRODUCT
AVGORDERS
byREGION
InsertInsert
UpdateUpdate
DeleteDelete
InsertInsert
UpdateUpdate
DeleteDelete
DeferredDeferredRefreshRefresh
FullFullRefreshRefresh
ORDERSORDERS
IncrementalIncrementalRefreshRefresh
For Oracle employees and authorized partners only. Do not distribute to third parties. 2009 Oracle Corporation Proprietary and Confidential
Core data warehouse featuresParallel execution
select c.cust_last_name
, sum(s.amount_sold)
from customers c, sales s
where c.cust_id = s.cust_id
group by c.cust_last_name;
Data on Disk Parallel Servers
scanscan
scanscan
scanscan
aggregateaggregate
Scanners
Coordinator
joinjoin
joinjoin
joinjoin
aggregateaggregate
aggregateaggregate
Joiners Aggregators
RAC aware!
-
5/22/2018 DW Implementation Boot Camp - Student Manual
14/118
For Oracle employees and authorized partners only. Do not distribute to third parties. 2009 Oracle Corporation Proprietary and Confidential
Core data warehouse featuresPartitioning
Core functionality Performance Manageability
Oracle8 Range partitioning
Global range indexes
Static partitionpruning
Basic maintenanceoperations: add,drop, exchange
Oracle8i Hash and compositerange-hash partitioning
Partition-wise joins
Dynamic pruning
Merge operation
Oracle9i List partitioning Global indexmaintenance
Oracle9iR2 Composite range-listpartitioning
Fast partition split
Oracle10g Global hash indexes Local Indexmaintenance
Oracle10gR2 1M partitions per table Multi-dimensionalpruning
Fast drop table
OracleDatabase 11g
More composite choicesREF PartitioningVirtual Column Partitioning
Interval PartitioningPartition Advisor
For Oracle employees and authorized partners only. Do not distribute to third parties. 2009 Oracle Corporation Proprietary and Confidential
Core data warehouse featuresData compression
Tables and indexes can be compressed Can be specified on a per-partition basis
Typical compression ratio 4:1
Requires more CPU to load data Decompression minimal resource cost
Compress for all DML operations
Less data on disk Requires less time to read
Little or no impact on query performance
Completely transparent
Especially applicable to data warehouses
4XCompression
For Oracle employees and authorized partners only. Do not distribute to third parties. 2009 Oracle Corporation Proprietary and Confidential
Core data warehouse featuresData compression real world examples
0
500
1000
1500
2000
2500
Table Scans
0
0.1
0.2
0.3
0.4
2.5x Faster
DML Performance
0
10
20
30
40
< 3% Overhead
3x Saving
Data Storage
For Oracle employees and authorized partners only. Do not distribute to third parties. 2009 Oracle Corporation Proprietary and Confidential
Core data warehouse featuresDatabase Resource Manager
High PrioritySales Analysis20 users
(DOP 10)
Medium PriorityAd Hoc Reports200 users
(DOP 4)
Low PriorityETL Jobs
200 users
(DOP 4)
Protect the system pro-actively Maximum number of concurrent operations
Control CPU and IO consumption based on Consumer Groups
-
5/22/2018 DW Implementation Boot Camp - Student Manual
15/118
For Oracle employees and authorized partners only. Do not distribute to third parties. 2009 Oracle Corporation Proprietary and Confidential
Core data warehouse featuresSQL Access Advisor
Based on workload
Materialized view advice
Index advice
Partition advice
For Oracle employees and authorized partners only. Do not distribute to third parties. 2009 Oracle Corporation Proprietary and Confidential
Core data warehouse featuresSQL Access Advisor how it works
Indexes Material izedviews
Materializedviews log
SQL AccessAdvisor
Hypothetical
SQL cache
FilterOptions
STS
CompleteWorkload
Partitionedobjects
Hashpartitions?
Intervalpartitions?
For Oracle employees and authorized partners only. Do not distribute to third parties. 2009 Oracle Corporation Proprietary and Confidential
Core data warehouse featuresOracle Warehouse Builder
Design Target WarehouseDesign Target Warehouse1
2 Define SourcesDefine Sources
Relational Flat Files Applications Mainframe
Map Source to TargetMap Source to Target3
Oracle or flat files
Generate CodeGenerate Code4
Instantiate WarehouseInstantiate Warehouse5
6 Extract & transform data
For Oracle employees and authorized partners only. Do not distribute to third parties. 2009 Oracle Corporation Proprietary and Confidential
Core data warehouse featuresOracle Warehouse Builder - Repository
Multi-user repository Single update, multi-read
Object locking
Synchronize
Metadata import / export &interchange Bridges to Oracle tools, ERwin,
PowerDesigner
Bridges to other metadata availablefrom Metadata Integration Technology
Incorporated
-
5/22/2018 DW Implementation Boot Camp - Student Manual
16/118
For Oracle employees and authorized partners only. Do not distribute to third parties. 2009 Oracle Corporation Proprietary and Confidential
Core data warehouse featuresOracle Data Integrator
Next Generation Architecture
E-LTE-LTLoadExtract
Transform Transform
Conventional ETL Architecture
Extract LoadTransform Performance
Heterogeneous E-LT
Productivity Declarative design
Flexibility Active integration design
Easily extendable Knowledge modules
For Oracle employees and authorized partners only. Do not distribute to third parties. 2009 Oracle Corporation Proprietary and Confidential
Core data warehouse featuresReal Application Clusters
QueryQueryQueryQueryETL ETL ETL ETL
Data Warehouse
Scale across cluster
For Oracle employees and authorized partners only. Do not distribute to third parties. 2009 Oracle Corporation Proprietary and Confidential
Core data warehouse featuresReal Application Clusters
QueryQueryETL ETL
Data Warehouse
Scale across cluster
Dedicate nodes tospecific work
For Oracle employees and authorized partners only. Do not distribute to third parties. 2009 Oracle Corporation Proprietary and Confidential
Core data warehouse featuresReal Application Clusters
QueryQueryETL ETL
Data Warehouse
Scale across cluster
Dedicate nodes tospecific work
Scale out whenworkload increases
Query
-
5/22/2018 DW Implementation Boot Camp - Student Manual
17/118
For Oracle employees and authorized partners only. Do not distribute to third parties. 2009 Oracle Corporation Proprietary and Confidential
Spreads database files acrossall available storage for optimalperformance and resourceutilization
Enables simple, online resourceallocation and providesautomatic data rebalancing
Up to 3-way mirroring
Can leverage external hardwareRAID mirror and parity protection
Core data warehouse featuresAutomatic Storage Management (ASM)
Oracle OLAP OptionEmbedded, manageable, enterprise-ready
A multidimensional calculation and aggregation engine Multidimensional data types: Cubes and dimensions
An OLAP API for cube definition and multidimensional queries
A SQL interface to OLAP Cubes and dimensions
Embedded in the Oracle Database Runs within Oracle instance on same server
OLAP cubes are stored in Oracle data f iles
OLAP metadata in the Oracle Data D ictionary
All key database functionality extends to OLAP Security: Object and data security
Scalability: Real Application Clusters
Availability: Backup-Recovery, Disaster Recovery, RAC
Administration: Enterprise Manager
For Oracle employees and authorized partners only. Do not distribute to third parties. 2009 Oracle Corporation Proprietary and Confidential
Why Oracle OLAP is so FastData Access Method
Data stored in dense arrays
Offset addressing no joins
More powerful analysis
Better performance
Time
Product
CD
ProfitDVD
Audio
Q1 Q2 Q3SF
West
USA
Market
Compare Profit this Quarter versus Last Quarter
What is a Products Profit Share of its Category?
For Oracle employees and authorized partners only. Do not distribute to third parties. 2009 Oracle Corporation Proprietary and Confidential
For Oracle employees and authorized partners only. Do not distribute to third parties. 2009 Oracle Corporation Proprietary and Confidential
Core data warehouse featuresCurrent adoption
0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100%
DB Res Mgr
RMAN
ASM
Read Only
VPD
MV Use
Compression
Parallel Exec
Partitioning
-
5/22/2018 DW Implementation Boot Camp - Student Manual
18/118
For Oracle employees and authorized partners only. Do not distribute to third parties. 2009 Oracle Corporation Proprietary and Confidential
Oracle 11g data
warehousing features
For Oracle employees and authorized partners only. Do not distribute to third parties. 2009 Oracle Corporation Proprietary and Confidential
Oracle 11g data warehouse featuresData Result Cache
Automatically caches results of queries, query blocks, or PL/SQLfunction calls
Cache is shared across statements and sessions on serv er
Significant speed up for read-only / read-mostly data
Full consistency and proper semantics
Cache refreshed when any underlying table updated
join
join
Table1 Group-by
join
Table2 Table3
Q2: Uses resulttransparentlycache
For Oracle employees and authorized partners only. Do not distribute to third parties. 2009 Oracle Corporation Proprietary and Confidential
Oracle 11g data warehouse featuresData Results Cache
Retail customer data (~50 GB) Concurrent users submitting queries randomly
Executive dashboard with 12 heavy analytical queries
Cache results only at in-line view level
12 queries run in random, different order 4 queries cached
Measure average, total response time for all users
447 s
267 s
186 s
No cache
334 s
201 s
141 s
Cache
25%
25%
24%
Improvement
8
4
2
# Users
For Oracle employees and authorized partners only. Do not distribute to third parties. 2009 Oracle Corporation Proprietary and Confidential
Oracle 11g data warehouse featuresSQL PIVOT/UNPIVOT
Rotate rows into columns and vice versa Common operation for crosstab queries
PROD QUARTER AMOUNT
Shoes Q1 2000
Shoes Q2 1000
Jeans Q1 1000
Jeans Q3 500
Jeans Q3 100Jeans Q4 1000
PROD Q1 Q2 Q3 Q4
Shoes 2000 1000 Null Null
Jeans 1000 Null 600 1000
SELECT * FROM tPIVOT (sum(amount)
FOR quarter inQ1,Q2,Q3,Q4);
-
5/22/2018 DW Implementation Boot Camp - Student Manual
19/118
For Oracle employees and authorized partners only. Do not distribute to third parties. 2009 Oracle Corporation Proprietary and Confidential
Oracle 11g data warehouse featuresAdditional features
Enhanced partitioning options
Advanced compression
OLAP-based materialized views
Priority-dependent maximum degree of parallelism inDatabase Resource Manager
For Oracle employees and authorized partners only. Do not distribute to third parties. 2009 Oracle Corporation Proprietary and Confidential
Oracle data warehousing
hardware
CustomCustom
Complete Flexibility
Any OS, any platform
Easy fit into acompanys ITstandards
Oracle DW Hardware SolutionsThe evolution of hardware infrastructure
For Oracle employees and authorized partners only. Do not distribute to third parties. 2009 Oracle Corporation Proprietary and Confidential
CustomCustom
Complete Flexibility
Any OS, any platform
Easy fit into acompanys ITstandards
Documented best-practiceconfigurations fordata warehousing
ReferenceConfigurations
ReferenceConfigurations
Oracle DW Hardware SolutionsThe evolution of hardware infrastructure
For Oracle employees and authorized partners only. Do not distribute to third parties. 2009 Oracle Corporation Proprietary and Confidential
-
5/22/2018 DW Implementation Boot Camp - Student Manual
20/118
CustomCustom
Complete Flexibility
Any OS, any platform
Easy fit into acompanys IT
standards
Documented best-practiceconfigurations fordata warehousing
Optimized
Warehouse
Optimized
Warehouse
Scalable systemspre-installed and pre-
configured: ready torun out-of-the-box
Reference
Configurations
Reference
Configurations
Oracle DW Hardware SolutionsThe evolution of hardware infrastructure
For Oracle employees and authorized partners only. Do not distribute to third parties. 2009 Oracle Corporation Proprietary and Confidential
CustomCustom
Complete Flexibility
Any OS, any platform
Easy fit into acompanys IT
standards
Documented best-practiceconfigurations fordata warehousing
Optimized
Warehouse
Optimized
Warehouse
Scalable systemspre-installed and pre-
configured: ready torun out-of-the-box
Highest performance
Pre-installed and pre-configured
Sold by Oracle
Reference
Configurations
Reference
ConfigurationsSun Oracle
Database
Machine
Sun Oracle
Database
Machine
Oracle DW Hardware SolutionsThe evolution of hardware infrastructure
For Oracle employees and authorized partners only. Do not distribute to third parties. 2009 Oracle Corporation Proprietary and Confidential
Build From Scratchwith Components
OWI ReferenceConfigurations
Take delivery of OWI
configuration
Weeks to Months
Pre-implementationSystem sizing
Acquisition ofcomponents
Installation andconfiguration
Acquisition ofcomponents
Installation andconfiguration
Testing andValidation
Testing andValidation
Weeks to Months
Sun OracleDatabase Machine
Faster deploymentLower Risk
< 1 Week
Database Machine BenefitsAccelerate Implementations & Lower Risk
For Oracle employees and authorized partners only. Do not distribute to third parties. 2009 Oracle Corporation Proprietary and Confidential
For Oracle employees and authorized partners only. Do not distribute to third parties. 2009 Oracle Corporation Proprietary and Confidential
Analytics and data
mining
-
5/22/2018 DW Implementation Boot Camp - Student Manual
21/118
SQL analytics
Bring Algorithms to the Data - Not Data tothe Algorithms
Analytic computationsdone in the database SQL analytics
Statistics
OLAP
Data Mining
Scalability
Security
Backup & Recovery
Simplicity
OLAP
Data Mining
Statistics
For Oracle employees and authorized partners only. Do not distribute to third parties. 2009 Oracle Corporation Proprietary and Confidential
Why Data Mining?Do you wish to ..
Identify factors most associated with a business problem?
Predict customer behavior ?
Predict or estimate a value ?
Find profiles of targeted people or items?
Segment a population ?
Determine important relationships/market baskets within thepopulation ?
Find fraudulent or rare events ?
(Attribute Importance)
(Classification)
(Regression)
(Decision Trees)
(Clustering)
(Associations)
(Anomaly Detection)
For Oracle employees and authorized partners only. Do not distribute to third parties. 2009 Oracle Corporation Proprietary and Confidential
Statistics & SQL Analytics
Ranking functions rank, dense_rank, cume_dist, percent_rank, ntile
Window Aggregate functions(moving and cumulative)
Avg, sum, min, max, count, variance, stddev,first_value, last_value
LAG/LEAD functions Direct inter-row reference using offsets
Reporting Aggregate functions Sum, avg, min, max, variance, stddev, count,
ratio_to_report
Statistical Aggregates Correlation, linear regression family, covariance
Linear regression Fitting of an ordinary-least-squares regression line
to a set of number pairs.
Frequently combined with the COVAR_POP,COVAR_SAMP, and CORR functions.
Descriptive Statistics average, standard deviation, variance, min, max, median (via
percentile_count), mode, group-by & roll-up
DBMS_STAT_FUNCS: summarizes numerical columns of atable and returns count, min, max, range, mean, stats_mode,variance, standard deviation, median, quantile values, +/- nsigma values, top/bottom 5 values
Correlations Pearsons correlation coefficients, Spearman's and Kendall's
(both nonparametric).
Cross Tabs Enhanced with % statistics: chi squared, phi coefficient,
Cramer's V, contingency coefficient, Cohen's kappa
Hypothesis Testing Student t-test , F-test, Binomial test, Wilcoxon Signed Ranks
test, Chi-square, Mann W hitney test, Kolmogorov-Smirnovtest, One-way ANOVA
Distribution Fitting Kolmogorov-Smirnov Test, Anderson-Darling Test, Chi-
Squared Test, Normal, Uniform, Weibull, Exponential
Pareto Analysis 80:20 rule, cumulative results table
For Oracle employees and authorized partners only. Do not distribute to third parties. 2009 Oracle Corporation Proprietary and Confidential
Oracle Data Mining OptionEmbedded, manageable, enterprise-ready
A complete data-mining solution 10+ data-mining algorithms: addresses every business scenario
Generalized Linear Models (Logistic & Multiple Regression)
Oracle Data Miner: Graphical user-interface for analysts
A Data Mining PL/SQL & Java API for developers
Embedded in the Oracle Database Runs within Oracle instance on same server
Data Mining algorithms execute directly on database t ables
Data Mining models and metadata stored within Oracle database
All key database functionality extends to Data Mining Scalability: Parallelism, Real Application Clusters
Security: Object and data security
Administration: Enterprise Manager Portability: All major platforms
For Oracle employees and authorized partners only. Do not distribute to third parties. 2009 Oracle Corporation Proprietary and Confidential
-
5/22/2018 DW Implementation Boot Camp - Student Manual
22/118
Oracle Data MiningOracle in-Database Mining Engine
Oracle Data Miner (GUI) Simplified, guided data mining
Spreadsheet Add-In forPredictive Analytics 1-click data mining from a
spreadsheet
PL/SQL API & Java (JDM) API Develop advanced analytical
applications
Wide range of algorithms
For Oracle employees and authorized partners only. Do not distribute to third parties. 2009 Oracle Corporation Proprietary and Confidential
For Oracle employees and authorized partners only. Do not distribute to third parties. 2009 Oracle Corporation Proprietary and Confidential
Oracle Business
Intelligence
OLTP & ODSSystems
DataWarehouseData Mart
Oracle Business Intelligence EEData warehouse components
For Oracle employees and authorized partners only. Do not distribute to third parties. 2009 Oracle Corporation Proprietary and Confidential
OLTP & ODSSystems
DataWarehouseData Mart
SAP, OraclePeopleSoft, Siebel,
Custom Apps
FilesExcelXML
BusinessProcess
Oracle Business Intelligence EEFull data foundation
For Oracle employees and authorized partners only. Do not distribute to third parties. 2009 Oracle Corporation Proprietary and Confidential
-
5/22/2018 DW Implementation Boot Camp - Student Manual
23/118
OLTP & ODSSystems
DataWarehouseData Mart
SAP, OraclePeopleSoft, Siebel,
Custom Apps
FilesExcelXML
BusinessProcess
Oracle BI Server
Common Enterprise Information Model
Oracle Business Intelligence EECommon information model
For Oracle employees and authorized partners only. Do not distribute to third parties. 2009 Oracle Corporation Proprietary and Confidential
OLTP & ODSSystems
DataWarehouseData Mart
SAP, OraclePeopleSoft, Siebel,
Custom Apps
FilesExcelXML
BusinessProcess
Oracle BI Server
Common Enterprise Information Model
Reporting&
Publishing
Ad hocAnalysis
ProactiveDetectionand Alerts
MicrosoftOffice
InteractiveDashboards
Advantages:
Consolidate and
standardize BI
tools
Unified BI
metadata and
infrastructure
Seamless BI userexperience
Intelligent
Caching Services
Data Federation
Intelligent Alerting
Oracle Business Intelligence EEComprehensive, integrated BI suite
For Oracle employees and authorized partners only. Do not distribute to third parties. 2009 Oracle Corporation Proprietary and Confidential
Oracle BI Enterprise Edition Plus
(OBIEE+)
OLTP & ODSSystems
DataWarehouseData Mart
SAP, OraclePeopleSoft, Siebel,
Custom Apps
FilesExcelXML
BusinessProcess
PerformanceManagementApplications
Insight
PerformanceAction
SetGoals
Plan
Monitor
Analyze
Report
Align
HyperionEssbase
Oracle BI Server
Common Enterprise Information Model
Reporting & Publishing
Financial Reporting
BI Publisher
Interactive Reporting
SQR Production Reporting
Web Analysis
Ad hocAnalysis
ProactiveDetectionand Alerts
MicrosoftOffice
InteractiveDashboards
Extend
ODBC
For Oracle employees and authorized partners only. Do not distribute to third parties. 2009 Oracle Corporation Proprietary and Confidential
Summary
Oracle has many features to enhance datawarehouse operations
Oracles data warehouse features areimplemented transparently to developers
Oracles data warehouse features areintegrated with the Oracle database
Oracle has evolved the hardwareinfrastructure of data warehouses topreconfigured optimized warehouses
Oracle offers a complete set of analytic andbusiness intelligence capabilities
For Oracle employees and authorized partners only. Do not distribute to third parties. 2009 Oracle Corporation Proprietary and Confidential
-
5/22/2018 DW Implementation Boot Camp - Student Manual
24/118
For Oracle employees and authorized partners only. Do not distribute to third parties. 2008 Oracle Corporation Proprietary and Confidential
For Oracle employees and authorized partners only. Do not distribute to third parties.
2008 Oracle Corporation Proprietary and Confidential
For Oracle employees and authorized partners only. Do not distribute to third parties. 2008 Oracle Corporation Proprietary and Confidential
For Oracle employees and authorized partners only. Do not distribute to third parties. 2008 Oracle Corporation Proprietary and Confidential
Title of Presentation
Presenters NamePresenters Title
Parallel execution
-
5/22/2018 DW Implementation Boot Camp - Student Manual
25/118
For Oracle employees and authorized partners only. Do not distribute to third parties. 2009 Oracle Corporation Proprietary and Confidential
Module Agenda
Parallel execution overview
Parallel execution setup
Parallel joins
Parallel execution with RAC
For Oracle employees and authorized partners only. Do not distribute to third parties. 2009 Oracle Corporation Proprietary and Confidential
Parallel execution
overview
What is parallel execution?
With parallel execution, a task is Broken into smaller sub-tasks, which are executed independently in
parallel, with the result joined at completion
For Oracle employees and authorized partners only. Do not distribute to third parties. 2009 Oracle Corporation Proprietary and Confidential
SELECT c.cust_last_name, sum(s.amount_sold)FROM customers c, sales s
WHERE c.cust_id = s.cust_idGROUP BY c.cust_last_name;
Data on Disk Parallel Servers
scanscan
scanscan
scanscan
aggregateaggregate
Coordinator
joinjoin
joinjoin
joinjoin
aggregateaggregate
aggregateaggregate
For Oracle employees and authorized partners only. Do not distribute to third parties. 2009 Oracle Corporation Proprietary and Confidential
SQL Parallel Execution
QC is theQC is the user session that initiates the parallel SQL statementuser session that initiates the parallel SQL statement& it will& it will distribute the work to parallel serversdistribute the work to parallel servers
Parallel serversParallel servers -- individual sessions that performindividual sessions that perform
work in parallel They are allocated from awork in parallel They are allocated from a pool ofpool ofglobally available parallel server processes andglobally available parallel server processes andassigned to a given operationassigned to a given operation
-
5/22/2018 DW Implementation Boot Camp - Student Manual
26/118
Dynamic parallelismAdvantages
Uses granules
Advantages Uses optimal queueing technique
Avoids convoy effect
Adaptable to changing conditions
Structural
Runtime
Allow partitioning based on business requirements
For Oracle employees and authorized partners only. Do not distribute to third parties. 2009 Oracle Corporation Proprietary and Confidential
Parallel executionParallel scans
Data is partitioned into granules(block range or partition) Dynamic Granules Computed
during runtime by the database
Each scan parallel process isassigned multiple granules
No two scanners ever contendfor the same granule
Granules are assigned so thatthe load is balanced across allparallel scan processes
PQ #2
PQ #3
PQ #1
Parallel ExecutionServers
Parallel executionIntra- and inter-operation parallelism
PQ 1
PQ 2
PQ 3
PQ 4
Sales tableFull ScanParallel Sort
A - C
D - F
G - M
N - O
PQ 5
PQ 6
PQ 7
PQ 8
S
T - V
W - Z
P - R
SELECT ... FROM product p, sales s where p.prod_id = s.prod_id ORDER BY p.prodname;
UserProcess
QueryCoordinator
Intra-Parallelism Intra-Parallelism
DOP = 8(Intra-Parallelism)
DOP = 16(Inter-Parallelism)
Inter-Parallelism
For Oracle employees and authorized partners only. Do not distribute to third parties. 2009 Oracle Corporation Proprietary and Confidential
Parallel executionParallel scans of multiple tables
SELECT c.cust_name, s.date,
s.amount
FROM sales s, customers c
WHERE s.cust_id= c.cust_id;
Querycoordinator
SALES
Table
CUSTOMERS
Table
-
5/22/2018 DW Implementation Boot Camp - Student Manual
27/118
-
5/22/2018 DW Implementation Boot Camp - Student Manual
28/118
For Oracle employees and authorized partners only. Do not distribute to third parties. 2009 Oracle Corporation Proprietary and Confidential
Parallel executionParallel scans of multiple tables
SELECT c.cust_name, s.date,
s.amount
FROM sales s, customers c
WHERE s.cust_id= c.cust_id;
Querycoordinator
P1 P2 P3 P4
SALES
Table
CUSTOMERS
Table
Assembly
Degree of Parallelism = number of parallel execution serversassociated with a parallel operation
Assigned for database, object, session or query DOP is calculated at run-time
Default DOP (when not specified) =CPU_COUNT xPARALLEL_THREADS_PER_CPU
DOP may be reduced depending on the availability of parallelexecution servers and system load
Database Resource Manager can throttle DOP per resourceconsumer groups
Parallel executionDegree of Parallelism (DOP)
DOP applies to intra-operation parallelism In inter-operation pallelism, parallel execution servers
can be twice the number specified by the DOP No more than two sets of parallel execution servers
can be used for one parallelized statement at a time
Parallel executionDegree of Parallelism (DOP) - continued
SQL Access Methods Table Scans
Fast Full Index Scans
Partitioned Index Range Scans Sorts, Aggregations, Set Operations
External Table access Joins
Nested Loop, Sort-Merge Hash, Star, Partition-Wise
DDL CTAS, CREATE INDEX, REBUILD INDEX
DML INSERT-SELECT, UPDATE, DELETE, MERGE
Parallel executionParallelizable operations
-
5/22/2018 DW Implementation Boot Camp - Student Manual
29/118
For Oracle employees and authorized partners only. Do not distribute to third parties. 2009 Oracle Corporation Proprietary and Confidential
Parallel execution setup
Parallel Execution Server Pool Server pool size is determined by PARALLEL_MIN_SERVERS
and PARALLEL_MAX_SERVERS Dynamic based on demand
Adaptive Multiuser and DOP PARALLEL_ADAPTIVE_MULTI_USER = TRUE (default) Throttles DOP to prevent overloading of server
PARALLEL_MIN_PERCENT is the minimum parallel executionservers requested for the operation
Default 0 Disabled Operation not executed if the MIN % of DOP servers not available
Parallel execution setupConfiguration
Enabling Parallel DML, DDL and QueryALTER SESSION (ENABLE | DISABLE | FORCE) PARALLEL
(DML | DDL | QUERY) (PARALLEL n);
FORCEPARALLEL overrides default DOP Hints overrides a forced DOP
Parallel Query & Parallel DDL is enabled if ANY ofbelow is TRUE
Object was created with PARALLEL clause ALTER SESSION [ENABLE|FORCE]PARALLEL Parallel hints are specified in the query
Parallel DML enabled if both are TRUE ALTER SESSION [ENABLE|FORCE]PARALLEL Object was created with PARALLEL clause OR Parallel
Hints are specified in DML
Parallel execution setupExecution
For Oracle employees and authorized partners only. Do not distribute to third parties. 2009 Oracle Corporation Proprietary and Confidential
Paralleljoins
-
5/22/2018 DW Implementation Boot Camp - Student Manual
30/118
Serial Join
scan scan
join
Products Sales
Sales.prod_id = products.prod_id
Scan of both tablesin serial
Compares (joins) allrecords from saleswith products
Basic Parallel Join
scan scan
join
scan scan
join
Sales.prod_id = products.prod_id
Scan of both tablesin parallel
Compares (joins) allrecords from saleswith products
Products Sales
Parallel Join with Broadcasting
scan scan
join
scan
join
Sales.prod_id = products.prod_id
Scan of both tablesin parallel/serial
Compares (joins) allrecords from saleswith products
Products Sales
Smaller table isbroadcast to all
processes
Parallel Join with Broadcasting
scan scan
join
scan
join
Sales.prod_id = products.prod_id
Scan of both tablesin parallel/serial
Compares (joins) allrecords from saleswith products
Smaller table isbroadcast to allprocesses
No data distributionfor large table
Products Sales
-
5/22/2018 DW Implementation Boot Camp - Student Manual
31/118
Parallel processesData Redistribution Methods
HASH Standard redistribution for parallel processing
Example JOIN, GROUP BY operations
KEY Ensures the logical clumping of KEY values
RANGE Used for combined GROUP BY and ORDER BY operations
BROADCAST Sends smaller data sets to all participating processes to
optimize data shipping
ROUND ROBIN Final operation to send data to requesting process
Variance LOCAL as RAC optimization
Partition-wise Join
scan scan
join
Sales.prod_id = products.prod_id
Products Sales
Scan of both tablesin serial
Compares (joins) allrecords from saleswith products
Equi-partitioning oftables will beleveraged
Partition-wise Parallel Join
scan scan
join
scan scan
join
Sales.prod_id = products.prod_id
Products Sales
Scan of both tablesin serial
Compares (joins) allrecords from saleswith products
Equi-partitioning oftables will beleveraged
No data distributionrequired
For Oracle employees and authorized partners only. Do not distribute to third parties. 2009 Oracle Corporation Proprietary and Confidential
Parallel execution with
RAC
-
5/22/2018 DW Implementation Boot Camp - Student Manual
32/118
Parallel Query
PQ 1
PQ 2
PQ 3
PQ 4
Sales tableFull ScanParallel Sort
A - C
D - F
G - M
N - O
PQ 5
PQ 6
PQ 7
PQ 8
S
T - V
W - Z
P - R
SELECT ... FROM product p, sales s where p.prod_id = s.prod_id ORDER BY p.prodname;
UserProcess
QueryCoordinator
Internode Parallel Query
PQ 1
PQ 2
PQ 3
PQ 4
Sales tableFull ScanParallel Sort
A - C
D - F
G - M
N - O
PQ 5
PQ 6
PQ 7
PQ 8
S
T - V
W - Z
P - R
SELECT ... FROM product p, sales s where p.prod_id = s.prod_id ORDER BY p.prodname;
UserProcess
QueryCoordinator
Node 1
Node 2
Many large-scale DWs have many concurrent jobs
Multiple small-to-medium size queries
Degree of parallelism < CPUs-per-node
With Oracle, queries will automatically run on a single node,eliminating traffic over the interconnect
RAC and Parallel Execution
Q1 Q2 Q4Q3
Q5 Q7Q6 Q8
Q9 Q12Q11Q10
Very large queries utilize all resources on thecluster
RAC and Parallel Execution
Large Query
-
5/22/2018 DW Implementation Boot Camp - Student Manual
33/118
RAC and parallel executionAllocating parallel execution servers
If no DOP has been set for the object, session orthrough a hint, RAC uses parallel servers from acrossall nodes.
Else, if DOP < 2 times the number of CPUs on anode, RAC will first attempt to use parallel executionprocesses on owner node
If DOP is too high or if enough local parallel serverprocess are not available, processes spread overmultiple nodes
P1 P1 P2 P2 P3 P3 P4 P4
Node 1 Node 2
PX1 PX2 PX3 PX4
Internode Parallel QuerySample optimizations
Classical co-located partition-wise join A partition-wise parallel join is used if the DOP is either equal to
the number of partitions, or an even divisor of the number ofpartitions
If there were 16 partitions, a partition-wise parallel join wouldwork with a DOP of 16, 8, 4 or 2
Example shown 2 nodes, 4 partitions on 2 tables, DOP 4
P1 P1 P2 P2 P3 P3 P4 P4
Node 1 Node 2
Internode Parallel QuerySample optimizations
Without partition-wise parallel join, data redistributionmust occur, which can stress interconnect Sample join of two tables, each with 4 partitions, on 2 nodes,
DOP 8
PX1 PX2 PX3 PX4 PX5 PX6 PX7 PX8
Parallel Operations with RACUsing services
PARALLEL_INSTANCE_GROUPS andINSTANCE_GROUPS nolonger used to controlparallelism across RACnodes
Services specify whichnodes can participate inparallel execution
Connection of a processto a service throughsql*net can contain theparallel execution
Define service thru OEM
or (DB and srvctl) thenadd to tnsnames.ora
-
5/22/2018 DW Implementation Boot Camp - Student Manual
34/118
Summary
Oracle uses dynamic parallelism
The degree of parallelism is dynamicallycalculated, based on several factors
Parallel partition-wise joins are theperformance goal
Parallelism is properly supported with RAC
Instance groups or services can limitparallelism to specific nodes with RAC
For Oracle employees and authorized partners only. Do not distribute to third parties. 2009 Oracle Corporation Proprietary and Confidential
For Oracle employees and authorized partners only. Do not distribute to third parties. 2008 Oracle Corporation Proprietary and Confidential
For Oracle employees and authorized partners only. Do not distribute to third parties.
2008 Oracle Corporation Proprietary and ConfidentialFor Oracle employees and authorized partners only. Do not distribute to third parties.
2008 Oracle Corporation Proprietary and Confidential
-
5/22/2018 DW Implementation Boot Camp - Student Manual
35/118
For Oracle employees and authorized partners only. Do not distribute to third parties. 2008 Oracle Corporation Proprietary and Confidential
Title of Presentation
Presenters NamePresenters Title
Partitioning
For Oracle employees and authorized partners only. Do not distribute to third parties. 2009 Oracle Corporation Proprietary and Confidential
Module Agenda
Partitioning benefits
Partitioning options
Index partitioning
Partition operations
For Oracle employees and authorized partners only. Do not distribute to third parties. 2009 Oracle Corporation Proprietary and Confidential
Partitioning benefits
Dividing a table or index into smaller portions,based on a partition key
Partitioning transparent to applications and users
PartitioningDefinition
For Oracle employees and authorized partners only. Do not distribute to third parties. 2009 Oracle Corporation Proprietary and Confidential
-
5/22/2018 DW Implementation Boot Camp - Student Manual
36/118
Partition pruning
Optimizer eliminates unneeded partitions for data access
Improves query performance by several orders (FTS .vs. FPS)
Prunes range or list partitioning for LIKE, equality, and IN-list
Prunes hash partitioning for equality and IN-list
Static and dynamic partition pruning
Partition-wise join
Used when two tables are joined together and when both the tables arepartitioned on the join key
Breaks the large join into smaller joins that occur between each of thepartitions
Significantly reduces response time and improves the use of both CPU andmemory resources
Partitioning benefitsPerformance
For Oracle employees and authorized partners only. Do not distribute to third parties. 2009 Oracle Corporation Proprietary and Confidential
Maintenance operations can address smallerpartitions rather than full table or index
Maintenance operations like DROP, SPLIT,COALESCE make it easy to adjust partitions
Partitions support a 'rolling window load or purgeprocess
Partitioning benefitsManageability
For Oracle employees and authorized partners only. Do not distribute to third parties. 2009 Oracle Corporation Proprietary and Confidential
Partitioned database objects provide partitionindependence
Important part of a high-availability strategy
Reduce scheduled downtime
Partitioning benefitsAvailability
For Oracle employees and authorized partners only. Do not distribute to third parties. 2009 Oracle Corporation Proprietary and Confidential
Partitions can be stored on different storage tiers
Archived partitions can be stored on cheaper storagedevices
Partitioning benefitsCost
For Oracle employees and authorized partners only. Do not distribute to third parties. 2009 Oracle Corporation Proprietary and Confidential
-
5/22/2018 DW Implementation Boot Camp - Student Manual
37/118
For Oracle employees and authorized partners only. Do not distribute to third parties. 2009 Oracle Corporation Proprietary and Confidential
Partition options
Range
Hash
List
Partitioning optionsClassic
For Oracle employees and authorized partners only. Do not distribute to third parties. 2009 Oracle Corporation Proprietary and Confidential
Partitioning optionsComposite partitioning
Table SALESRANGE(order_date)-RANGE(ship_date)
Jan 2006
... ...
Feb 2006 Mar 2006 Jan 2007
... ...
... ...
...
...
...
...
Jan2006
Feb2006
Jan2007
Data ispartitioned alongtwo dimensions(A,B)
A distinct valuepair for the twodimensionsuniquelydetermines thetarget partitioning
For Oracle employees and authorized partners only. Do not distribute to third parties. 2009 Oracle Corporation Proprietary and Confidential
Partitioning optionsComposite partitioning
Table SALESRANGE(order_date)-RANGE(ship_date)
Jan 2006
... ...
Feb 2006 Mar 2006 Jan 2007
... ...
... ...
...
...
...
...
Jan2006
Feb2006
May2006
Mar 2006
All records withorder_date inMarch 2006ANDship_date inMay 2006
May2006
For Oracle employees and authorized partners only. Do not distribute to third parties. 2009 Oracle Corporation Proprietary and Confidential
-
5/22/2018 DW Implementation Boot Camp - Student Manual
38/118
Composite partitioning options
11g11g11gList
8i9iR211gRange
HashListRange
Partitioning optionsComposite partitioning
For Oracle employees and authorized partners only. Do not distribute to third parties. 2009 Oracle Corporation Proprietary and Confidential
Partitioning optionsInterval partitioning
Extension to range partitioning
Full automation for equi-sized range partitions
Partitions are created automatically as data arrives No need to create new partitions
Local indexes are created and maintained as well
For Oracle employees and authorized partners only. Do not distribute to third parties. 2009 Oracle Corporation Proprietary and Confidential
Table SALES
Jan 2006
... ...
Feb 2006 Mar 2006 Jan 2007 Oct 2009 Nov 2009
...
Interval partitioned table can have classical range andautomated interval section Automated new partition management plus full partition
maintenance capabilities:Best of both worlds
Partitioning optionsInterval partitioning
For Oracle employees and authorized partners only. Do not distribute to third parties. 2009 Oracle Corporation Proprietary and Confidential
Table SALES
Jan 2006
... ...
Feb 2006 Mar 2006 Jan 2007 Oct 2009 Nov 2009
...
Interval partitioned table can have classical range andautomated interval section Automated new partition management plus full partition
maintenance capabilities: Best of both worlds
MERGE and move old partitions for ILM
Range partition section
Partitioning optionsInterval partitioning
For Oracle employees and authorized partners only. Do not distribute to third parties. 2009 Oracle Corporation Proprietary and Confidential
-
5/22/2018 DW Implementation Boot Camp - Student Manual
39/118
Table SALES
Jan 2006 Feb 2006 Mar 2006 Jan 2007 Oct 2009 Nov 2009
Interval partitioned table can have classical range andautomated interval section Automated new partition management plus full partition
maintenance capabilities:Best of both worlds
MERGE and move old partitions for ILM
Range partition section
... ... ...
Interval partition section
VALUES ('13-NOV-2009')Insert new data- Automatic segment creation
Partitioning optionsInterval partitioning
For Oracle employees and authorized partners only. Do not distribute to third parties. 2009 Oracle Corporation Proprietary and Confidential
Table SALES
...
2005 Q1 2006 Oct 2006
Range partitioned tables can be extended into intervalpartitioned tables Simple metadata command
Investment protection
Q2 2006
...
Partitioning optionsInterval partitioning
For Oracle employees and authorized partners only. Do not distribute to third parties. 2009 Oracle Corporation Proprietary and Confidential
Table SALES
...
2005 Q1 2006 Oct 2006Q2 2006
ALTER TABLE sales (order_date DATE, ...)SET INTERVAL(NUMTOYMINTERVAL(1,'month');
Range partitioned tables can be extended into intervalpartitioned tables Simple metadata command
Investment protection
...
New monthlyInterval partitions
Old range partition table
...
Partitioning optionsInterval partitioning
For Oracle employees and authorized partners only. Do not distribute to third parties. 2009 Oracle Corporation Proprietary and Confidential
Related tables benefit from same partitioning strategy Sample 3NF order entry data model
Redundant storage of the same information solvesthis problem Data overhead
Maintenance overhead
Business Problem
Solution Oracle Database 11gintroduces REF Partitioning
Child table inherits the partitioning strategy of parent tablethrough PK-FK relationship
Intuitive modelling Enhanced performance and manageability
Partitioning optionsREF partitioning
For Oracle employees and authorized partners only. Do not distribute to third parties. 2009 Oracle Corporation Proprietary and Confidential
-
5/22/2018 DW Implementation Boot Camp - Student Manual
40/118
Table ORDERS
Jan 2006
... ...
Feb 2006
Table LINEITEMS
Jan 2006
... ...
Feb 2006
RANGE(order_date)
Primary key order_id
RANGE(order_date)
Foreign key order_id
Redundant storage of order_date
Redundant maintenance
Partitioning optionsBefore REF partitioning
For Oracle employees and authorized partners only. Do not distribute to third parties. 2009 Oracle Corporation Proprietary and Confidential
Table ORDERS
Jan 2006
... ...
Feb 2006
Table LINEITEMS
Jan 2006
... ...
Feb 2006
RANGE(order_date)
Primary key order_id
RANGE(order_date)
Foreign key order_id
RANGE(order_date)
Foreign key order_id
PARTITION BY REFERENCE Partitioning key inherited through
PK-FK relationship
Partitioning optionsWith REF partitioning
For Oracle employees and authorized partners only. Do not distribute to third parties. 2009 Oracle Corporation Proprietary and Confidential
Want to create extended schema attributes are fullyderived and dependent on existing common data
Redundant storage or extended view definitions aresolving this problem today Requires additional maintenance and creates overhead
Business Problem
Solution Oracle Database 11g introduces virtual columns
Purely virtual, meta-data only
Treated as real columns except no DML Virtual columns can have statistics
Virtual columns are eligible as partitioning key Enhanced performance and manageability
Partitioning optionsVirtual column-based partitioning
For Oracle employees and authorized partners only. Do not distribute to third parties. 2009 Oracle Corporation Proprietary and Confidential
ORDERS
ORDER_ID ORDER_DATE CUSTOMER_ID...---------- ----------- ----------- --9834-US-14 12-JAN-2007 659208300-EU-97 14-FEB-2007 396543886-EU-02 16-JAN-2007 45292566-US-94 19-JAN-2007 153273699-US-63 02-FEB-2007 18733
REGION requires no storage
Partition by ORDER_DATE, REGION
REGION AS (SUBSTR(ORDER_ID,6,2))------
USEUEUUSUS
Partitioning optionsVirtual column-based partitioning
For Oracle employees and authorized partners only. Do not distribute to third parties. 2009 Oracle Corporation Proprietary and Confidential
-
5/22/2018 DW Implementation Boot Camp - Student Manual
41/118
For Oracle employees and authorized partners only. Do not distribute to third parties. 2009 Oracle Corporation Proprietary and Confidential
Index partitioning
Created as equi-partition with theunderlying table; created with theLOCAL attribute
the indexpartition key ranges andtable partition key ranges areidentical
Managed automatically by server toensure equi-partitioning
Easier administration
Enables parallelism of index scans
Reduction of scheduled downtime
Only one index partition affectedby maintenance operation on anunderlying table partitionPartition
Q1
Partition
Q2
Partition
Q3Partition
Q4
SALES07 Table
IndexPartition
IndexPartition
IndexPartition
IndexPartition
CREATE INDEX order_idxON sales07 (order_no)
LOCAL(PARTITION INDEX_Q1,PARTITION INDEX_Q2 ,PARTITION INDEX_Q3,PARTITION INDEX_Q4);
CREATE INDEX order_idxON sales07 (order_no)
LOCAL(PARTITION INDEX_Q1,PARTITION INDEX_Q2 ,PARTITION INDEX_Q3,PARTITION INDEX_Q4);
Local partitioned indexes
For Oracle employees and authorized partners only. Do not distribute to third parties. 2009 Oracle Corporation Proprietary and Confidential
41
Not equi-partitioned with theunderlying table
The index is partitioned by the indexkey, not by the table partition key
OLTP applications may getbetter performance becausethey minimize the number ofindex probes
Harder to manage than localindexes
Can be unique or non-uniquePartitionQ1
Partition
Q2
Partition
Q3Partition
Q4
SALES TablePartitioned by Quarter
CREATE INDEX product_idxON sales07 (product_id)
PARTITION BY RANGE(product_id)GLOBAL(PARTITION PIDX1 VALUES LESS THAN (1000),PARTITION PIDX2 VALUES LESS THAN (2000),PARTITION PIDX3 VALUES LESS THAN (MAXVALUE));
CREATE INDEX product_idxON sales07 (product_id)
PARTITION BY RANGE(product_id)GLOBAL(PARTITION PIDX1 VALUES LESS THAN (1000),PARTITION PIDX2 VALUES LESS THAN (2000),PARTITION PIDX3 VALUES LESS THAN (MAXVALUE));
Index PartitionProducts
0-999
Index PartitionProducts
1000-1999
Index PartitionProducts
2000 +
Global partitioned indexes
For Oracle employees and authorized partners only. Do not distribute to third parties. 2009 Oracle Corporation Proprietary and Confidential
All index data within one schemaobject
Treated the same as global indexes
Administration advantages gainedwhen table is partitioned are lost ifassociated indexes are notpartitioned
Why? If recovering a partition of a table,it will only be usable once the entire non-partitioned Index is recovered and usable
Non-Partitioned Index
Partition
Q1
Partition
Q2
Partition
Q3Partition
Q4
SALES07 Table
CREATE INDEX sales07_idxON sales97 (sales_date);
CREATE INDEX sales07_idxON sales97 (sales_date);
Non-partitioned indexes
For Oracle employees and authorized partners only. Do not distribute to third parties. 2009 Oracle Corporation Proprietary and Confidential
-
5/22/2018 DW Implementation Boot Camp - Student Manual
42/118
For Oracle employees and authorized partners only. Do not distribute to third parties. 2009 Oracle Corporation Proprietary and Confidential
Partition operations
Partition-wise join
scan scan
join
Sales.prod_id = products.prod_id
Products Sales
Scan of both tablesin serial
Compares (joins) allrows from sales withproducts
Equi-partitioning oftables will beleveraged
For Oracle employees and authorized partners only. Do not distribute to third parties. 2009 Oracle Corporation Proprietary and Confidential
Partition-wise parallel join
scan scan
join
scan scan
join
Sales.prod_id = products.prod_id
Products Sales
Scan of both tablesin serial
Compares (joins) allrecords from saleswith products
Equi-partitioning oftables will beleveraged
No data distributionrequired with RAC
For Oracle employees and authorized partners only. Do not distribute to third parties. 2009 Oracle Corporation Proprietary and Confidential
Rolling window operations
Q106 Q206 Q306 Q406
Order Table(partitioned by quarter)
Drop
Other data & queries not affected
Q107
Add
For Oracle employees and authorized partners only. Do not distribute to third parties. 2009 Oracle Corporation Proprietary and Confidential
-
5/22/2018 DW Implementation Boot Camp - Student Manual
43/118
EXCHANGE PARTITION
The basic technique of bulk-loading new data into atemporary load table, which is then indexed, analyzed,and then published all at once to end-users using theEXCHANGE PARTITION operation
Should be the default load technique for all large tables ina data warehouse
For Oracle employees and authorized partners only. Do not distribute to third parties. 2009 Oracle Corporation Proprietary and Confidential
22-Feb
2008
23-Feb
2008
24-Feb
2008
(empty)
Composite-partitioned
table TXN
2. Bulk
Loads
5. EXCHANGE PARTITION
3. Create
Indexes
4. Table
/ Col Stats
EXCHANGE PARTITION
1. Create
Temp Table
For Oracle employees and authorized partners only. Do not distribute to third parties. 2009 Oracle Corporation Proprietary and Confidential
25-Feb
2008
Hash-partitioned
table TXN_TEMP
Summary
Partitioning improves performance andmanageability
Oracle supports many types of partitioning,including range, list, hash, interval, virtualcolumn and reference partitions
Oracle supports composite partitions
Performance goal is partition-wise joins
Indexes can be local or global
Partition exchange can be used to add
data to a warehouse with very highavailability
For Oracle employees and authorized partners only. Do not distribute to third parties. 2009 Oracle Corporation Proprietary and Confidential
For Oracle employees and authorized partners only. Do not distribute to third parties. 2008 Oracle Corporation Proprietary and Confidential
-
5/22/2018 DW Implementation Boot Camp - Student Manual
44/118
For Oracle employees and authorized partners only. Do not distribute to third parties.
2008 Oracle Corporation Proprietary and ConfidentialFor Oracle employees and authorized partners only. Do not distribute to third parties.
2008 Oracle Corporation Proprietary and Confidential
For Oracle employees and authorized partners only. Do not distribute to third parties. 2008 Oracle Corporation Proprietary and Confidential
Title of Presentation
Presenters NamePresenters Title
Result Caches
For Oracle employees and authorized partners only. Do not distribute to third parties. 2009 Oracle Corporation Proprietary and Confidential
Module Agenda
SQL Result Cache
PL/SQL Function Cache
-
5/22/2018 DW Implementation Boot Camp - Student Manual
45/118
For Oracle employees and authorized partners only. Do not distribute to third parties. 2009 Oracle Corporation Proprietary and Confidential
SQL Result Cache
Analyze data across large data sets Reporting
Forecasting & trend analysis
Data mining
Use parallel execution for good performance
Result Very I/O intensive workload direct reads from disk
Memory is less important
PGA memory more sensitive based on sort or aggregations
SQL Result CacheData warehouse workloads
For Oracle employees and authorized partners only. Do not distribute to third parties. 2009 Oracle Corporation Proprietary and Confidential
SQL Result CacheData warehouse sample query
For Oracle employees and authorized partners only. Do not distribute to third parties. 2009 Oracle Corporation Proprietary and Confidential
Accesses many rows
Returns few rows
select p.prod_category, sum(s.amount_sold) revenuefrom products p, sales swhere s.prod_id = p.prod_idand s.time_idbetween to_date('01-JAN-2006','dd-MON-yyyy')and to_date('31-DEC-2006','dd-MON-yyyy')
group by rollup (p.prod_category)
SQL Result CacheWhat it does
For Oracle employees and authorized partners only. Do not distribute to third parties. 2009 Oracle Corporation Proprietary and Confidential
Caches results of queries, query blocks, orPL/SQL function calls
Read consistency is enforced DML/DDL against dependent database objects invalidates
cache
Bind variables parameterize cached result withvariable values
-
5/22/2018 DW Implementation Boot Camp - Student Manual
46/118
SQL Result CacheHow it works
For Oracle employees and authorized partners only. Do not distribute to third parties. 2009 Oracle Corporation Proprietary and Confidential
Table 1
Table 2 Table 3
join
join
Group by
query 1executes
SQL Result CacheHow it works
For Oracle employees and authorized partners only. Do not distribute to third parties. 2009 Oracle Corporation Proprietary and Confidential
Table 1
Table 2 Table 3
join
join
Group by
Table 1
Table 2 Table 3
join
join
Group bycachedresult
result iscached
SQL Result CacheHow it works
For Oracle employees and authorized partners only. Do not distribute to third parties. 2009 Oracle Corporation Proprietary and Confidential
Table 1
Table 2 Table 3
join
join
Group bycachedresult
Table 5 Table 6
join
Table 4
join
Group by
joinQuery 2 uses cachedresult transparently
RESULT_CACHE_MODE initialization parameter MANUAL, use hints to populate and use
FORCE, queries will use cache without hint
RESULT_CACHE_MAX_SIZE initialization parameter Default is dependent on other memory settings
(0.25% of MEMORY_TARGET or 0.5% of SGA_TARGETor 1% of SHARED_POOL_SIZE)
0 disables result cache
Never > 75% of shared pool (built-in restriction)
/*+ RESULT_CACHE */ hint in queries
SQL Result CacheEnabling SQL Result Cache
-
5/22/2018 DW Implementation Boot Camp - Student Manual
47/118
SQL Result CacheSQL hint example
Use RESULT_CACHE hint
select /*+ RESULT_CACHE */ p.prod_category, sum(s.amount_sold) revenuefrom products p, sales swhere s.prod_id = p.prod_idand s.time_idbetween to_date('01-JAN-2006','dd-MON-yyyy')and to_date('31-DEC-2006','dd-MON-yyyy')
group by rollup (p.prod_category)
SQL Result CacheView hint example
Use RESULT_CACHE hint in view definitionselect prod_subcategory, revenuefrom(select /*+ RESULT_CACHE */ p.prod_category, p.prod_subcategory, sum(s.amount_sold) revenuefrom products p, sales swhere s.prod_id = p.prod_idand s.time_idbetween to_date('01-JAN-2006','dd-MON-yyyy')and to_date('31-DEC-2006','dd-MON-yyyy')
group by rollup (p.prod_category, p.prod_subcategory))where prod_category = 'Women'/
Similar query for different prod_category can use cached result
SQL Result CacheExecution plan for query using result cache
------------------------------------------------------------------| Id | Operation | Name |------------------------------------------------------------------| 0 | SELECT STATEMENT | || 1 | RESULT CACHE | fz6cm4jbpcwh48wcyk60m7qypu || 2 | SORT GROUP BY ROLLUP | ||* 3 | HASH JOIN | || 4 | PARTITION RANGE ITERATOR| ||* 5 | TABLE ACCESS FULL | SALES || 6 | VIEW | index$_join$_001 ||* 7 | HASH JOIN | || 8 | INDEX FAST FULL SCAN | PRODUCTS_PK || 9 | INDEX FAST FULL SCAN | PRODUCTS_PROD_CAT_IX |------------------------------------------------------------------
Depends... based on Query repetitiveness
Query execution times
DML activity (cache invalidation frequency)
Remember data warehouse workload Query may run 30 minutes
Query may return 5 rows
Query served from result cache would take a split second
SQL Result CacheOpportunity for benefit
-
5/22/2018 DW Implementation Boot Camp - Student Manual
48/118
Use DBMS_RESULT_CACHE package to manage
Use V$RESULT_CACHE_* views to monitor
SQL Result CacheManagement
Result cache is disabled for queries containing Temporary or dictionary tables
Non-deterministic PL/SQL functions
Sequence CURRVAL and NEXTVAL
SQL functions current_date, sysdate, sys_guid, etc.
Result cache for distributed queries Set RESULT_CACHE_REMOTE_EXPIRATION > 0
0 means distributed queries are not cached
Default is 0
DML/DDL on remote database will not expire cached results
SQL Result CacheRestrictions
Result cache does not automatically release memory Grows until maximum size is reached
DBMS_RESULT_CACHE.FLUSH purges memory
Bind variables Cached result is parameterized with variable values
Cached results can only be found for same variable values
Cached result will not be built if Query is built on a non-current version of data (read consistency
enforcement)
Current session has outstanding transaction on table(s) in query
Flashback queries can be cached
SQL Result CacheCaveats
Retail customer data (~50 GB)
Concurrent users submitting queries randomly Executive dashboard application with 12 heavy analytical queries
Cache results only at in-line view level 12 queries run in random, different order
4 queries benefiting from the cache
Measure average, total response time for all users
# Users No cache Cache Im provem ent
2 186 s 141 s 24%
4 267 s 201 s 25%
8 447 s 334 s 25%
SQL Result CacheInternal benchmark
-
5/22/2018 DW Implementation Boot Camp - Student Manual
49/118
For Oracle employees and authorized partners only. Do not distribute to third parties. 2009 Oracle Corporation Proprietary and Confidential
PL/SQL Function Cache
Calculate a complex derived metric like the ratio ofthe highest median income grouped by state to thelowest median income grouped by state over thewhole population
You need a PL/SQL function
Like the results stored in the SQL Result Cache, thedata changes fairly slowly but the query is repeatedfairly often
PL/SQL Function CacheThe challenge
function f2 return t1%rowtype
is...
beginselect a, m into r1.a, r1.b from ...;
select a, m into r2.a, r2.b from ...;
r.a := r1.a + r2.a;r.b := r1.b + r2.b;return r;
end f2;
~ 2,000 milliseconds for each new call
PL/SQL Function CacheFunction implementation
function f2 return t1%rowtyperesult_cache relies_on(t1, t2)
is...
beginselect a, m into r1.a, r1.b from t1;
select a, m into r2.a, r2.b from t2;
r.a := r1.a + r2.a;r.b := r1.b + r2.b;return r;
end f2;
~ 0 milliseconds for each new call
PL/SQL Function CacheUsing the function cache
-
5/22/2018 DW Implementation Boot Camp - Student Manual
50/118
Both are cross-session and RAC interoperable Set RESULT_CACHE_MAX_SIZE > 0 on all instances, or = 0
on all instances
Memory pool is instance-specific
Both build on the same infrastructure Same RESULT_CACHE_MAX_SIZE initialization parameters
Same DBMS_RESULT_CACHE management package
Same V$RESULT_CACHE_*performance views
PL/SQL Function CacheSimilarities with SQL Result Cache
Summary
The Result Cache saves results of queries,reducing overhead needed on repeateduse
Result cache is invalidated if there is DMLor DDL on cached tables
Result Cache can be used to satisfy part ofa query
PL/SQL Function Cache can cachefunction results
For Oracle employees and authorized partners only. Do not distribute to third parties. 2009 Oracle Corporation Proprietary and Confidential
For Oracle employees and authorized partners only. Do not distribute to third parties. 2008 Oracle Corporation Proprietary and Confidential
For Oracle employees and authorized partners only. Do not distribute to third parties.
2008 Oracle Corporation Proprietary and Confidential
-
5/22/2018 DW Implementation Boot Camp - Student Manual
51/118
For Oracle employees and authorized partners only. Do not distribute to third parties. 2008 Oracle Corporation Proprietary and Confidential
For Oracle employees and authorized partners only. Do not distribute to third parties. 2008 Oracle Corporation Proprietary and Confidential
Title of Presentation
Presenters NamePresenters Title
Oracle OLAP
For Oracle employees and authorized partners only. Do not distribute to third parties. 2009 Oracle Corporation Proprietary and Confidential
Module Agenda
Oracle OLAP Option
Cube Organized Materialized Views
Querying OLAP Cubes
For Oracle employees and authorized partners only. Do not distribute to third parties. 2009 Oracle Corporation Proprietary and Confidential
Oracle OLAP Option
-
5/22/2018 DW Implementation Boot Camp - Student Manual
52/118
A full featured multidimensional OLAPserver
Excellent query performance for ad-hoc /unpredictable query
Enhances the analytic content of businessintelligence application
Fast