Oracle Grid Computing Grid Control Storage Databases Application

Oracle Grid Computing Grid Control Storage Databases Application Servers

George Demarest Senior Director Oracle 10 g Oracle Corporation

Evolution of IT Mainframe Departmental Servers Client-Server Internet Computing

“Today’s IT shops are criminally inefficient, running assets at a small fraction of their capacity. ” Frank E. Gillett Forrester Research, October 2002

Problems with IT Ÿ Islands of computation – – Configured for peak loads Limited scalability Availability < 99. x% Fragmented security Ÿ Escalating costs Ÿ Affects all businesses

What is Grid Computing?

Growing Grid Momentum

Grid versus Grid Academic Grid Computing Ÿ Computation-centric Ÿ Coarse data management Ÿ Address specialized computing problems Ÿ 1% of Fortune 1000 for 1% of their computing tasks Enterprise Grid Computing Ÿ Business alignment Ÿ Flexible data management Ÿ Addresses all enterprise computing requirements Ÿ 100% of Fortune 1000 for 99% of computing tasks

V I D E O The Grid Experts Speak

Grid Computing Defined Coordinated use of many servers and storage acting as unified computing resource

Reducing IT Costs Oracle Grid Large Dedicated Server Ÿ Ÿ Expensive costly components High incremental costs Single point of failure Enterprise service at high cost Ÿ Ÿ Low cost modular components Low incremental costs No single point of failure Enterprise service at low cost

V I D E O Customers Discuss Grid Computing

Introducing

Oracle Grid Computing Model • • • Pooling Virtualization & Provisioning Load Balancing Quality of Service Automation

You Can Start Small … Ÿ Easy installation – 1 CD, 20 minutes Ÿ Clone “known good” configurations Ÿ Automated application provisioning Ÿ Fully configured out of the box Ÿ Automated patch management

Self-Managing Database Ÿ Built-in intelligent infrastructure Ÿ Automatic SQL tuning Ÿ Automatic storage management Ÿ Automatic database diagnostic monitor

Automatic Storage Management

Oracle Real Application Clusters

Database and Application Server Clusters: Proven Ÿ Thousands of customers Ÿ Running on all platforms

Multi-tier Failover Notification (Fa. N) Ÿ Oracle Application Server becomes more RAC aware – Overcomes delays from cross-tier application failovers RAC Failover AS Detection Total Downtime Without Fa. N < 8 secs 15 mins > 15 mins With Fa. N < 8 secs < 4 secs < 12 secs

Unified Data Provisioning

Quality of Service Ÿ Continuous availability Ÿ Record-setting performance Ÿ Low cost scalability

High Availability Ÿ Faster disk-based backup and recovery Ÿ Flashback recovery Ÿ Data Guard Ÿ Fast failover of all applications – Seconds for total recovery

New World: Disk-Based Data Recovery 1980’s - 200 MB Ÿ Disk economics approaching tape Ÿ Disk is better than tape – 1000 x increase Random access to any data Ÿ Re-architected recovery strategy – backup and recover just the changes Ÿ Backup and Recovery goes from hours to minutes 2000’s - 200 GB

Flashback Error Correction Database Customer Ÿ Flashback Recovery at all levels Ÿ Flashback Database – Restores the whole database to some past time Ÿ Flashback Table – Order restores rows in a set of tables Ÿ Flashback Rows – restores rows

Oracle Database 10 g Flashback Database Ÿ Like a “Rewind” button for the Database – A new strategy for data recovery Ÿ Super efficient – Restores just changed data Ÿ It’s fast – << Recover in minutes, not hours Ÿ It’s easy – A single command restore operation “Show me the database as it was Tuesday night”

Scale Out for Grids Ÿ Cluster ALL services – – HTTP, Cache, Java, LDAP, Portal OLTP, business intelligence, content management Ÿ Grow in small increments Ÿ Create Ultra Large Databases

Database Performance 1, 184, 893 World Record TPC-C 28% Faster than Oracle 9 i Linux Cluster World Record on Linux 1, 008, 144 541, 764 423, 414 136, 111 Transaction November Processing 4, www. tpc. org. As of December 2003: Sixteen-node 1, 008, 144. 49 HP Integrity rx 5670 tpm. C, $8. 33/tpm. C, server cluster, available each with 12/31/03. 4 Itanium® 2 Database 1. 5 GHz 10 g Enterprise As As of of September 5, 5, 2003: Council Oracle(TPC), Database 10 g Enterprise Edition, HP HP 8, Integrity Superdome, 824, 164. 53 tpm. C, $8. 28/tpm. C, available 12/31/03. Oracle NEC Express 5800/1320 Xd, Oracle Edition, Database HP 9000 10 g Enterprise Superdome Source: processors, Transaction 1, 184, 893. 38 Processing tpm. C, Council (TPC), available www. tpc. org April 30, 2004. HP Superdome server with HP 64 10 g Itanium® 2 1. 5 GHz processors, HP-UX 11 iv 2, Edition, Enterprise 521, 440. 53 Server, 541, 673 tpm. C, $11. 77/tpm. C, $5. 52/tpm. C, $10. 69/tpm. C, available February 12/31/03. 15, Oracle 9 i HP Integrity Database rx 5670, Enterprise (Oracle Edition, Database 9000 Standard Superdome Edition), Enterprise 136, 110. 98 Server, tpm. C, 423, 414 $4. 09/tpm. C, $15. 64/tpm. C, available December available 31, 12/31/03 2003. 1, 008, 144. 49 tpm. C, $8. 33/tpm. C, available December 31, 2003 Source: Transaction Processing Council (TPC), www. tpc. org

Automation Ÿ Managing application service levels Ÿ Managing groups of systems Ÿ Managing security centrally

V I D E O Why Manageability Is Important

Enterprise Manager Grid Control Ÿ Monitor and manage Ÿ Grid-wide view Ÿ End-to-end Manage from a Browser Ÿ Top-to-bottom EM 2 Go Ÿ From anywhere … or a PDA

Service Level Management Monitor End-user Experience Availability Performance Monitor Database Click-to-SQL Drilldowns Monitor Application Click-to-EJB J 2 EE Activity

Manage Groups as One Ÿ Single-view management and monitoring Ÿ Standardize policies Applications – – – Configuration Performance Security Ÿ Automate processes Sets of Systems

Security & Identity Management Ÿ Centrally manage user identities Ÿ Centrally grant and revoke access privileges Ÿ Provide single sign-on access to applications Ÿ Share security policies across the infrastructure

Next Steps… Consolidate Standardize Automate

Consolidate Standardize Automate “The typical enterprise, with mainframe, Unix and Windows deployments, could save between 8. 5% & 10. 5% of the data center budget by implementing standardization, automation and consolidation. ” Gartner Research July 2003

Consolidate Standardize Automate “We wanted to gain economies of scale. ” Gretchen Mc. Coy Senior Vice President, VISA International

Consolidate Standardize Automate “Oracle proved you can run complex software economically on Linux. ” Mark Rizzo Chief Information Technology Architect, Electronic Arts

Consolidate Standardize Automate “We can manage more services, database, application servers in complex environments. ” Mary Martin Johnston-Turner Wientjes Summit Strategies, August 2003 Group Manager, Atos Origin

Oracle Data Center Ÿ Oracle University grid Ÿ Outsourcing grid Grid Control Storage Databases Application Servers

V I D E O Customers Discuss Oracle 10 g

Complete, integrated grid infrastructure

“The best thing about the Grid is that it is unstoppable” June 21, 2001 2

…Like Linux

For More Information On the Internet oracle. com/grid 1: 1 Consultations: oracle. com/consultations