1. Exadata Database Machine Morana Kobal Butković Senior Sales Consultant Oracle Hrvatska

2. Exadata Goals Ideal Oracle Database Platform
Best Machine for Data Warehousing
Best Machine for OLTP
Best Machine for Database Consolidation
Unique Architecture Makes it
Fast and cost efficient
Oracle has set the bar extremely high for Exadata. We are designing a single hardware platform that is best for any class of Database Workloads.
The hardware is identical for all workload types since high compute capacity, high network bandwidth, low latency, high IO throughput, and Flash help any workload.
The software features vary by workload. For example Oracle has many Warehouse specific software features such as Bitmap indexing, integrated OLAP and Data Mining, etc.
The Unique capabilities of Exadata are highlighted in the presentation by being underlined in Red

3. Exadata in the Marketplace
Launched in Fall 2008
Seeing rapid adoption in all geographies and industries

4. Agenda
Hardware Architecture
Key Technologies
Consolidation & Protection

5. The Products Exadata Storage Server & Database Machine
Storage Product Optimized for Oracle Database
Extreme I/O and SQL Processing performance
Combination of hardware and software
Exadata Storage Server Software
Exadata Database Machine
Pre-Configured High Performance
Balanced performance configuration
Takes the guess work out of building an Oracle deployment
Exadata Storage Server Software
Oracle Database 11.2

6. Exadata Storage Server Building Block
Hardware by Sun
Software by Oracle
Uses high performance components
12 disks GB 15K RPM SAS 2.0, or 2TB 7200 RPM SATA
2 Xeon quad-core processors with PCI 2.0
Dual ported 40 Gb/sec InfiniBand
4 96 GB PCI Flash Cards
Runs at full disk and flash bandwidth

7. Exadata Hardware Architecture
Scalable Grid of industry standard servers for Compute and Storage
Eliminates long-standing tradeoff between Scalability, Availability, Cost
Database Grid
Storage Grid
8 compute servers (1U)
64 Intel cores
14 storage servers (2U)
112 Intel cores in storage
100 TB SAS disk, or 336 TB SATA disk
5 TB PCI Flash
Data mirrored across storage servers
Oracle believes that the Scaleable Grid architecture is the architecture of the future. It eliminates the long-standing tradeoff between high-end hardware platforms that are scaleable (up to a point) and available, but also very high cost due to low volume.
Sometimes people think that because there is so much compute and storage in a database machine it must consume huge amounts of power. In fact the power consumption is not large. Maximum Power usage of a Full Rack Database Machine is 14KW, typical is 9.8 KW. A single high end SMP platform without storage or switches can consume well over 20 KW.
InfiniBand Network
Redundant 40Gb/s switches
Unified server & storage net

8. Database Server Hardware
Dual-redundant, hot-swappable power supplies
72 GB DRAM (18 x 4GB)
ILOM
4 x 2.5” 146GB Disk Drives
4 x 1GbE Interfaces
2 Quad-Core Intel® Xeon® E5540 Processors
InfiniBand QDR (40Gb/s) dual port card
Installed Software:
Oracle Enterprise Linux
Oracle Database 11.2 Software
Drivers
Disk Controller HBA with 512M battery backed cache

9. Start Small and Grow Field Upgradeable
Exadata allows easy expansion within and between racks.
A quarter rack can be upgraded to a half rack. A half rack can be upgraded to a full rack.
Two half racks can also be connected together to form the equivalent of a full rack. This is sometimes useful in data centers with weight or heat density restrictions.
Once a full rack is deployed it can be increased in half rack size increments. For example a full rack grows to 1.5 racks, then to 2 racks, then to 2.5 racks.
Equipment can be mixed across hardware generations. For example a half rack can be grown to a full rack using the next generation of servers to fill out the rack.
Quarter Rack
Half Rack
Full Rack
Balanced Incremental Scaling for OLTP and DW

10. Scales to 8 Racks by Just Adding Cables Full Bandwidth and Redundancy
Exadata has much more compute, storage, and interconnect capacity than other systems on the market. An 8 rack Exadata system is equivalent to at least 30 racks of leading conventional products, and costs much less.
The user data capacity is much larger after compression.
Scales beyond 8 racks by using external InfiniBand switches 10

11. Exadata Database Machine Product Family
Quarter Rack
Half Rack
Full Rack
2-8 Full Racks
Database Servers 2 4 8
16-64
Exadata Storage Servers 3 7 14
28-112
Total Disk Capacity SAS
21 TB
50 TB
100 TB
200 – 800TB
Total Disk Capacity SATA
70 TB
168 TB
336 TB
336 – 2688TB
User Data (uncompressed SAS)
6 TB
14 TB
28 TB
56 – 224 TB
I/O Throughput (disks SAS)
4.5 GB/sec
10.5 GB/sec
21 GB/sec
GB/sec
I/O Throughput (flash)
11 GB/sec
25 GB/sec
50 GB/sec
GB/sec
Flash IOPS
225,000
500,000
1,000,000
1M – 8M
Racks 1
2-8

12. Standardized and Simple to Deploy
All Database Machines are the same
Delivered Tested and Ready-to-Run
Highly Optimized
Highly Supportable
No unique configuration issues
Identical to config used by Oracle Engineering
Runs existing OLTP and DW applications
Full 30 years of Oracle DB capabilities
No Exadata certification required
Leverages Oracle ecosystem
Skills, knowledge base, people, partners
Ready- to-Run
Eliminates the complexity of deploying a high performance database system.
Database machines are tested in the factory and delivered ready to run. Because all database machines are the same, their characteristics and operations are well known and understood by Oracle field engineers and support. Each customer will not need to diagnose and resolve unique issues that only occur on their configuration. Performance tuning, and stress testing performed at Oracle is done on the exact same configuration that the customer has ensuring better performance and higher quality.
Applications do not need to be certified against Exadata. Applications that are certified with Oracle Database 11.2 RAC will run against Exadata. Very few applications need to certify the storage subsystem underneath a database, and Exadata fundamentally is the Oracle Database with a very fast storage subsystem.
Deploy in Days, Not Months

13. Agenda
Hardware Architecture
Key Technologies
Consolidation & Protection

14. Keys to Speed and Cost Advantage
Exadata Hybrid Columnar Compression
Exadata Intelligent Storage Grid
Exadata Smart Flash Cache

15. Exadata Intelligent Storage Grid Most Scalable Data Processing
Data Intensive processing runs in Exadata Storage Grid
Filter rows and columns as data streams from disks
Example: How much product X sold last quarter
Exadata Storage Reads 10TB from disk
Exadata Storage Filters rows by Product & Date
Sends 100GB of matching data to DB Servers
Scale-out storage parallelizes execution and removes bottlenecks
Exadata storage servers scan the data in parallel removing all central bottlenecks.
Traditional storage has bottlenecks at the back-end disk loops, caches, controllers, front-end channels, and SAN links.
In traditional storage, receiving Terabytes of data into the DB server and filtering it consumes large amount of CPU in the database hosts.

16. Simple Query Example SUM Exadata Storage Grid Oracle Database Grid
What were my sales yesterday?
Optimizer Chooses Partitions to Access
Exadata Storage Grid
Oracle Database Grid
Scan compressed blocks in partitions
Retrieve sales amounts for Sept 24
Select sum(sales)
where
Date=’24-Sept’
SUM
10 TB scanned
100 GB returned to servers 16

17. Exadata Intelligent Storage
Exadata Intelligent Storage Grid
Exadata storage servers also run more complex operations in storage
Join filtering
Incremental backup filtering
I/O prioritization
Storage Indexing
Database level security
Offloaded scans on encrypted data
Data Mining Model Scoring
Smart File Creation
10x reduction in data sent to DB servers is common

18. Exadata is Smart Storage
Storage Server is smart storage, not a DB node
Storage remains an independent tier
Database Servers
Perform complex database processing such as joins, aggregation, etc.
Exadata Storage Servers
Search tables and indexes filtering out data that is not relevant to a query
Cells serve data to multiple databases enabling OLTP and consolidation
Simplicity, and robustness of storage appliance
Compute and Memory Intensive Processing
Data Intensive Processing

19. Exadata Hybrid Columnar Compression
Data is organized and compressed by column
Dramatically better compression
Speed Optimized Query Mode for Data Warehousing
10X compression typical
Runs faster because of Exadata offload!
Space Optimized Archival Mode for infrequently accessed data
15X to 50X compression typical
Query
Faster and Simpler
Backup, DR, Caching, Reorg, Clone
Benefits Multiply

20. Exadata Hybrid Columnar Compression How it works
Compression Unit
Tables are organized into sets of a few thousand rows
Compression Units (CUs)
Within CU, data is organized by column, then compressed
Column organization brings similar values close together, enhancing compression
Useful for data that is bulk loaded and queried
Update activity is light
Exadata servers offload filtering, projection, etc. for scans on compressed data
Return compressed blocks to database so buffer cache benefits from compression
Column 2
Column 1
Column 3
Reduces
Table Size
4x to 40x
4x to 50x Reduction

21. Compression Ratio of Real-World Data
Compression Ratio varies by customer and table
Trials were run on largest table at 10 large companies
Average Query Compression ratio was 13x
On top of Oracle’s already highly efficient format
This slide shows the compression ratio achieved on the largest table in Exadata trials at 10 ultra large companies. We chose to report the ratio for the largest table since that gives a much better idea of the effective size reduction than averaging ratios across large and small tables. Compressing small tables generally does not provide a huge net space savings. These tables generally have many billions of rows.
Compression ratio means how many times smaller the final data was. For example a 1TB table that compresses to 100GB would have a 10x compression ratio.
Note that some vendors quote compression ratios starting from highly verbose flat file formats. This allows them to inflate their compression ratios by 50% to 100%.
Oracle compression ratios are based on data that is already in Oracle’s highly efficient data format that uses compact variable length representations for all data types. Customers typically find that Oracle’s uncompressed data representation uses 30% to 50% less space than other database vendors. 21

22. Hybrid Columnar Comparisons
Uncompressed
Uncompressed
Pure Columnar Cliff
OLTP Compress
OLTP
Hybrid
Hybrid
Pure Column
Pure Column
Table Size
Scan Time
Row Lookup Time
Exadata Hybrid Columnar Compression is a second generation columnar technology combining the best of row and column formats
Best compression – matching full columnar
Excellent scan time – Compression provides 10x speedup
After 10x I/O reduction, most queries become CPU bound
Good single row lookup – no full columnar “cliff”
Full columnar format falls off a cliff for single row lookup operations
Must do 100 I/Os for a 100 column table to “stitch” a row back together - 100x slowdown.
Often missing any secondary indexing capabilities. If full scan is only access method then slowdown is 1000x or more.
Many systems that are considered Warehouses have extensive row lookups and trickle feeds.
Pure columnar is a niche technology that is not useful in OLTP, and is useful in only some warehouses.
Hybrid columnar gains almost all the compression advantages of full columnar compression, as well as the CPU reductions from column at a time processing. Because of the high compression factors achieved by hybrid and full columnar technologies, queries often become CPU bound instead of I/O bound, or the full data set fits in memory. In those cases there is NO advantage to full columnar.
93% statement is based on 1000 GB scan which becomes 100 GB scan with hybrid and 30 GB Scan with full columnar (assuming 33% of columns referenced). Hybrid gives 90% reduction compared to 97% reduction for full columnar. 90/97=93%. 22

23. Exadata Smart Flash Cache Breaks the Disk Random I/O Bottleneck
Trade-off between traditional disks drives and Flash
Disk drives are cheap, high capacity but low I/Os per second
Flash is expensive, lower capacity but can support tens of thousands of I/Os per second
Ideal Solution - Exadata Smart Flash Cache
Keep most data on disk for low cost
Transparently move hot data to flash
Use flash cards instead of flash disks to avoid disk controller limitations
Flash cards in Exadata storage
High bandwidth, low latency interconnect
300 I/O per Sec
Tens of Thousands of I/O’s per Second

24. Sun FlashFire in Exadata Sun Flash Accelerator F20 PCIe Card
Cell cache on the storage cells
Write-though cache, transparently used to accelerate reads
4 x Cards (384GB/cell) used to create a cache on the cell
Database Machine has 5 TB of flash storage
Able to pull 3.6GB/sec total bandwidth from each storage cell, for Full Rack:
50GB/sec total from flash
21GB/sec from SAS disk, 12Gb/sec SATA disk
Up to 50,000 Disk IOPS SAS or Up to 20,000 Disk IOPS SATA
Up to 1,000,000 Flash IOPS

25. Exadata Smart Flash Cache
Performance
Use PCIe cards instead of SSDs to avoid slow disk interface
Capacity
Efficient Compression increases effective performance and capacity by up to 10X
Smart Caching
Caches data intelligently to maximize Flash usage for frequently read data
Automatically skips caching of infrequently read objects or avoid caching data that will not fit in the cache
Database awareness enables caching only data likely to be accessed again
User can further optimize caching policies by specifying whether or not to cache specific database objects
Understands different types of I/Os from database
Skips caching I/Os to mirror copies
Skips caching backups
I/Os due to ASM rebalance operations not cached
Skips caching data pump I/O
Skips caching tablespace formatting
Resistant to table scans
Control File Reads and Writes are cached
File header reads and writes are cached
Data Blocks and Index blocks are cached
DBA can enforce that an object is kept in flash cache
ALTER TABLE calldetail STORAGE (CELL_FLASH_CACHE KEEP)
Can be set like other storage clause values
At table level, partition level, during creation time etc.

26. Exadata is Architected for Flash
Traditional storage arrays now offer optional flash disks
Exadata Flash Architecture
Scale-Out Storage
No bottlenecks to scaling flash I/O
InfiniBand
Highest throughput, lowest latency
Intelligent Storage
Key to using full flash bandwidth
Even InfiniBand can’t send 50GB/sec
PCI Flash
Avoids disk controller bottlenecks.
Cards in storage enable HA, RAC
Compression
Multiply flash capacity 10x
Also multiplies data scan rates
Flash Cache
Speed of flash, cost of disk
Optionally specify table placement

27. Exadata Storage Index Transparent I/O Elimination with No Overhead
Table
Index
Exadata Storage Indexes maintain summary information about table data in memory
Store MIN and MAX values of columns
Typically one index entry for every MB of disk
Eliminates disk I/Os if MIN and MAX can never match “where” clause of a query
Completely automatic and transparent A B C D 1 3 5 8
Min B = 1
Max B =5
Min B = 3
Max B =8
Select * from Table where B< Only first set of rows can match 27

28. Benefits Multiply Example
10 TB of user data
Normally 10 TB of IO
1 TB
with 10x compression
100 GB
with partition pruning
Seconds
On Exadata
50 GB
with Storage Indexes
Scan 50GB in Flash
10 GB returned after Exadata Filtering
Data is 10x Smaller, Scan is 2000x faster

29. Agenda
Hardware Architecture
Key Technologies
Consolidation & Protection

30. Unified Hardware, Specialized Software
Exadata enables a single hardware architecture for all database needs
Massively parallel hardware, InfiniBand, & Flash for all DB applications and workloads
Enables Strategic building block approach to IT
Domain specialization is in software, not hardware
Analytics
OLAP, Statistics, Spatial, Data Mining, etc.
Warehousing
Flexible Partitioning, Bitmap Indexing, Join indexing, Materialized Views, Result Cache
Data
Relational, XML, Objects, Secure Files
OLTP, Security, HA
OLAP
ETL
Others argue that a specialized hardware or software platform is needed for each application or class of work.
Biggest driver of ongoing cost is system sprawl
Hardware configurations are isolated and customized to specific applications
Special systems for high performance, fault tolerance, low cost, data warehousing
Database machine enables database consolidation
Better performance than high-end systems
With fault tolerance built in
At high volume price
Runs any combination of workloads with extreme performance
Warehouse oriented bulk data processing
OLTP oriented random updates
Real time BI against transactional data
Data Mining

31. Platform for Database Consolidation
ERP
CRM
Warehouse
Data Mart HR
Consolidation is key to reducing costs
Administration, hardware, software, data center
Many databases can be consolidated on Exadata
Multiple small databases within a node
Large databases can span nodes using RAC
Exadata serves as farm/cloud for databases
Exadata delivers performance for complex workloads that mix OLTP and DW
Complex OLTP with batch and reporting
Complex Warehousing with thousands of users
Multiple databases running different applications

32. Consolidate Database Storage
ERP
CRM
Warehouse
Data Mart HR
Exadata and ASM allow all storage servers to be shared across databases
Shared Configuration
Advanced ASM data striping spreads every database across all storage servers
Eliminates hot-spots and captive unused space
Full storage grid performance available to all databases
Predictable Performance
Exadata I/O resource manager prioritizes I/Os to ensure predictable performance
At user, job, application, or database level
No need for isolated storage islands

33. Consolidate Database Servers
Many databases can run on Database Machine servers
Shared Configuration
Applications connect to a database service that runs on one or more database servers
Services can grow, shrink, & move dynamically
Large databases can span nodes using RAC
Multiple small databases can run on a single node
Predictable performance
Instance caging provides predictable CPU resources when multiple databases run on the same node
Restricts a database to subset of processors
ERP
CRM
Warehouse
Data
Mart HR

35. Resources Oracle.com: http://www.oracle.com/exadata
Oracle Exadata Technology Portal on OTN: