Sybase Adaptive Server Enterprise Product Direction Jan Gipe – IPG Product Mgmt jgipe@sybase

1. Sybase Adaptive Server EnterpriseProduct DirectionJan Gipe – IPG Product Mgmtjgipe@sybase.com

2. Agenda • ASE Scorecard • ASE Overview and Product Direction • Competitive Advantages • Replication Overview and Product Direction

3. Addressing customer needs Faster delivery of features utilizing in-market releases, 12.5.0.1 and 12.5.0.3, 12.5.1 Delivery of features that customers want (improved operational scalability, Linux support, DBXray visual diagnostic tool, LDAP authentication) Working with partners Working with Veritas on support for cluster solutions Collaboration with HP around Itanium New and improved tools with BMC and Lecco Tech– DB Expert, DBXray Commitment to product Continued investment to maintain leadership position in specific areas Superior performance on 4-way Intel/Linux Technology innovation First to support XML natively in server; new support for Xquery and web services Complete range of availability solutions, including new support for snapshot replication Executing on 2 year roadmap to meet next round of customer requirements ASE in Sybase’s strategic vision ASE is a key component in Sybase’s strategic direction in Real-time services Move from database to data management platform ASE Scorecard

4. Application Development Rapidly Evolving Business Models Multiple Development Options (XML, Java, SQL) New Legislative Requirements Data Management & Operations Higher Data Availability More Data Online Growing Data Sizes Platform & Infrastructure Cost Pressures Business Continuity Scarcity of Skills Time-to-Market Budget Enterprise Data ManagementToday’sIT Focus – Total Cost of Operations Impacts on TCO: Operational Constraints:

5. Storage & Access Volumes Operational Capacity Very Large Operational Gap RDBMS Large Small 1980 2000 1990 The Operational Gap More Data, More Transactions, Limited Resources

6. Sybase ASE TCO Leader in Enterprise Data Management “Based on our analysis of thousands of applications, the total life-cycle cost of most RDBMS applications will be, on average, more than 15 percent lower when running on Sybase ASE versus the equivalent offering from Oracle Corporation” - Standish Group, 2003

7. Sybase ASE Inherent TCO Advantages Application Development Costs 20% Web Architecture Service Oriented Architecture Web Services & XML Database Operational Costs Integrated Data Delivery 60% Self- Management Virtualized Data Services 20% Lower Cost Platforms Provisioning Grid Computing ASE 12.5.1 Key Technologies Future ASE Technologies TCO Components

8. Sybase ASE 12.5 Family Overview 12.5.0.3 12.5.0.1 12.5 • Security • Secure Sockets Layer • Row based access controls • Performance • New Page Sizes: 4K, 8K & 16K pages • Dynamic Tuning • Availability • Additional HA Cluster Support • Compressed Backups • Enhanced QuiesceDB • Application Development • Fast EJB’s • XML Indexing • Access to External File System • Availability • ASE Replicator (built-in Replication) • Enhanced HA • Migration • Migration Tools • Application Development • JVM & XML Enhancements • Administration • DB-Xray • Sampling based Statistics • Migration • PowerTransfer Migration Utility • Availability • Sun Clusters: Active/Passive

9. Sybase ASE 12.5.1 – Design Focus Higher Operational Scalability Reduced Human Intervention Lowering the Cost of Applications • Simplified administration tasks • Simpler skill sets • Fewer user errors • Reduced planned downtime Improved Developer Productivity • Reduced learning curve • Less coding • Easily expose data as a service Improved System Performance Lower Hardware Costs • More Throughput with Less H/W • Manage more data and transactions • Transition to Low Cost hardware while maintaining performance Lowest TCO

10. Sybase ASE 12.5.1 – Design Focus Higher Operational Scalability Reduced Human Intervention Lowering the Cost of Applications • Administration Improvements • DBA & Developer Tools • Self Management • Security & Availability Improved Developer Productivity • Web Services Advancements • Enhanced XML • Derived Tables • Globalization Enhancements Improved System Performance Lower Hardware Costs • TempDB Enhancements • I/O Enhancements • Faster Maintenance Operations • Linux Performance Lowest TCO

11. Transportable Databases Quick copies for reporting or online backups Ship pre-configured databases Minimize downtime during application re-provisioning Dynamic Cache Management Reduces planned downtime and increases availability during application tuning Simplified Installer Easy configuration Administration Improvements Operational Scalability Reduce Planned Downtime Improve DBA Productivity

12. Tools for DBAs and Developers Operational Scalability • DBXray • Monitor server health & bottlenecks with drill down capabilities • DB Expert • Analyze, predict, optimize and manage database performance • SQL Expert • Tune application performance DBXray Improve Application Performance DBExpert

13. Job Scheduler Schedule consistency checks, index management and reports Templated Backups –predefined templates for scheduling backups Templated Resource Management – predefined templates to monitor and dynamically adjust resources like connections and locks Automatic DB Space Management Automatically grow databases and logs with thresholds Reduced unplanned downtime with self healing Automatic Garbage Collection Cleanup unused space automatically Self Management Operational Scalability Reduce Planned Downtime Reduce Human Intervention

14. Security & Availability Operational Scalability • Simplified Security Administration • Centralized User Management through LDAP • Single Login/Password • Availability • Dynamic Listener Support • Add capacity on demand • Recover quickly from listener attacks • Enhanced Cluster support for highly available multi-node configurations • Sun Cluster 3.0, Veritas VCS 3.5 Login Authenticate LDAP Server Improve Application Availability

15. Multiple TempDB Reduced contention for tempdb by assigning applications or users to different tempdbs Optimized TempDB Logging Increased throughput with reduced log space contention Optimized TempDB I/O Performance of TempDB operations (select into) improved with optimized I/O TempDB Enhancements: Reduce Contention System Performance Improvement % Contention 15 10 Contention on catalog 5 0 5 10 15 20 25 Number of clients 1 Temp DB 2 Temp DB Higher Throughput Higher Application Performance

16. Select Into Enhancements Optimized Logging reduces the number of I/O Operations Use of large memory buffers reduce the number of I/O Operations Create/Load/Alter Database Enhancements Significant increase in performance with the use of Large I/O buffers Higher degree of Parallelism in operating on a single or multiple devices for these large operations ASE 12.0 ASE 12.5.1 I/O Enhancements System Performance Improvement Create Database 12000 10000 8000 6000 4000 2000 0 Total Time Higher Throughput Higher Application Performance

17. Parallel Recovery Faster Recovery at fail-over improves Availability Reduces downtime for large environments Parallel Check Points Reduces Recovery Time through more frequent check points and more I/O at each check point Sampling Based Statistics Reduced maintenance time for updating statistics for very large data sets Faster Maintenance Operations System Performance Improvement System Recovery Time 100 80 Recovery Time Percentage 60 40 20 0 1 2 3 Number of Recovery Tasks Reduced Planned Downtime

18. High Performance on Linux System Performance Improvement Sybase/HP joint performance benchmark results. • Best 1-4 CPU OLTP performance • Best response time • Low resource requirement (4GB vs SQL Server 32GB memory) System Recovery Time 70000 60000 50000 40000 30000 20000 10000 Trans/Min 0 1 2 3 4 Number of CPUs 4 2.0Ghz Intel Xeon processors and 4GB of memory on Proliant Systems

19. XML Enhancements XML Storage and Retrieval Fast XML Indexes XML-SQL Interoperability XPATH and XQUERY Web Services Support ASE as a producer : Publish data as Web service ASE Web Services Producer Input: SOAP Input: JDBC Output: SOAP Output: JDBC Any SOAP client ASE Support for Open Standards Lower Cost of Applications Faster Application Development Less Custom Code

20. Message Services Web Services RelationalData Internal SQL Query or External XQuery XML Data Translation to Relational Form XMLInput XML Processing in ASE – Today RelationalData XMLOutput • Support both native handling and “shredding” of XML data • Performance enhancements and support for multi-document queries • XQUERY Support • Global indexing, providing enhanced query performance; upgrades to our Full-Text Search service to be more XML-aware, as well as optimal handling of the XQUERY syntax • Support for “for XML” clause that is SQLX complaint • Representation of SQL result sets in XML, including mappings for SQL names, datatypes, and values. • A new “XML” datatype for SQL

21. Native XML Support Store and Retrieve XML documents natively – Schema independent and dynamic Patent pending fast XML Indexes (as efficient as B-Tree for relational data) – provides a very high performance for querying Self Defined indexes – no user input necessary SQL and XML duality Complete interoperability between XML and SQL at the language and storage level Select … from … FORXML will send the results in XML format Select … Xmlparse(xmldoc) will parse the XML document XPATH and XQUERY Support Fully support XQUERY 1.0 and XPATH 2.0 standard XQUERY/XPATH for querying XML data (like SQL for relational data) Support wildcards and functions for complex querying XML Views/Mapping & Transformation Can easily map XML data as SQL and SQL data as XML Can apply XML transformations to stored XML document as well as the XML view of the SQL data XML Support in ASE – Today

22. Expose logic inside ASE as a Web Service ASE as a Web Service Requestor/Orchestrator Use ASE to access a Web Service and use the result along with other data Expose logic inside ASE as a Web Service Details Access T-SQL (stored procedures, user defined functions) as web methods Persistent database connections via sessions Support SSL RPC/encoded web services support Access T-SQL as web methods Generate dynamic WSDL ASE Web Services Producer Input: SOAP Input: JDBC Output: SOAP Output: JDBC Any SOAP client ASE Web Services Support in ASE 12.5.1

23. Support for Derived Tables Better support for ISV/VAR applications Enhanced application portability to ASE platform ANSI SQL 2002 Compliance Unicode Enhancements Enhanced support for applications dealing with global data Chinese Character Set – GB18030 Support Compliance with Chinese Regulations Improved ISV support & Globalization Lower Cost of Applications Less Custom Code

24. Sybase ASE 12.5.1Takes Care of Business Sybase ASE is the most cost-effective data management platform for business-critical computing. • Operational Scalability – reducing costs of people, systems and time • System Performance – handling more data, transactions, and applications with existing resources • High Performance on Linux – lower cost platforms for business critical tasks • XML and Advanced Web Services – faster application development

25. ASE Future Roadmap 2004-2005 Higher Operational Scalability • Self Management – Zero Admin • Security, Auditing & Archiving • Availability Clusters • VLDB Enhancements • Partitions Improved System Performance • New Optimizer • New Execution Engine • Scalability Clusters Lowering the Cost of Applications • XQUERY Support • XML Transformations • Managed Services • Real Time Services Lowest TCO

26. Job Scheduler Job Scheduler Resource Management Module Resource Management Module Self Management – Today and Future Self Management Subsystems Polling ASE Object Fault Fault Configuration Event Space Dynamic Configuration Dynamic Configuration ASE 12.5.1 Future ASE

27. Event management Deliver alerts triggered by system or user-defined events to Job Scheduler Space management Database shrink capability Auto data de-fragmentation and re-org for data and index Object management Index usage/tracking across boots Configuration management Reduce number of configuration knobs Heuristics for automatic growth and shrinkage of configuration values Query tuning and optimizer performance Auto update statistics Self-learning, self-tuning optimizer Fault management Run data integrity checks when cycles are available to continue from last time Eliminate faults that can be managed within ASE and schedule tasks to manage them High availability Indices made online as soon as rebuilt Index reorg or other utilities do not block data access Empirical monitoring & feedback system Policy definition/enforcement Self Management – Future Directions

28. Segment-based Partitions Partitioned tables Range, Hash partition support Add/Drop one or more partitions Clustered/Non-Clustered Indexes Global and local indexes on partitions Update statistics on one or all partitions Truncate partition Optimizer and Execution support for partitioned tables New Limits More than 256 devices More than 32 segments Remove 30 char limit on names Object Level Recovery Recover database objects from database dumps Make database consistency checks an offline activity Row locked system catalogs VLDB Enhancements – Future

29. Improved System PerformanceNew Optimizer Efficient Query Optimization

30. New Platform: Benchmark Results Execution Time I/O Count Query 1 Query 2 Query 3 Query 1 Query 2 Query 3 Database size – 630 MB Query 1: Join with expressions and no indexes (infinite improvement) Query 2: Vector aggregation with Group By (exec time improvement – 186%; I/O savings – 83%) Query 3: Multiple scalar aggregation (exec time improvement – 571%; I/O savings – 93%)

31. HA System Establish Companion Node2 Node1 ASE1 ASE2 Shared Disk Storage Replicate Users/Logins Disk Disk Sybase Cluster Solutions - Current2-Node Active/Active ASE Solution • Although the disk is available to both nodes, the databases are not shared by the nodes – only at failure time are the disks from the failed node accessed by the companion node • 2 Node Cluster Solutions • Each node is a fail-over candidate for the other node • Both nodes run can active applications

32. Sybase Cluster Solutions – CurrentActive/Passive, Multi-node Solutions • Multi-Node Cluster Solutions • Multiple nodes are joined in a cluster, eg., Egenera Bladeframe systems • N+1 Node Cluster Availability • Typically, one node is passive and is designated the fail-over candidate for any of the other nodes • Protects against a single failure among N+1 nodes • Cluster software mechanisms also provide multi-node (Active/Passive configuration) availability • Cluster software detects failure of an active node and fails over ASE and disks to the fail-over candidate as configured previously (ASE server instance is started on the fail-over node) Cluster Software Active Nodes Passive Node Shared Disk System

33. Availability Clusters • N Node Cluster Availability • All nodes may run applications (some may be intentionally left idle to ensure service levels after failures) • Any node is a fail-over candidate for any of the other nodes • Protects against multiple failures As mission critical systems cannot tolerate even a few minutes or seconds of latency at fail-over time (typically required for recovery of the failed databases), a new approach is required • Each node in the cluster will run an Active ASE Server instance • The databases and disks are accessible *and* available to each ASE server instance (the databases are essentially shared by all nodes) • When one of the nodes fail, any of the other nodes can seamlessly takeover the databases of the failed node • Since each database is always available to all the nodes, no recovery needs to be performed on the database at fail-over time • Sybase’s future solution for availability will provide near-instantaneous fail-over with minimal latency and downtime • Initially focused on 2-4 node clusters, will be enhanced to larger number of nodes in the future

34. Shared Disk System Availability Clusters – Future • As the disks and databases are shared among the nodes, client applications can be migrated to any node seamlessly and continue processing

35. Clusters for Scalability • To achieve true scalability across a number of nodes, the application or data needs to be partitioned or mirrored on multiple nodes • Application Aware Partitioning • Data is partitioned across independent database servers and disks • Each server instance is completely independent and unaware of other data partitions • Client Application (or the mid-tier application server) is aware of the partitioning and directs the query/transaction to the corresponding node/server instance • If a query or transaction needs to access data on multiple nodes, the mid-tier application server coordinates the distribution among the nodes and the subsequent join of the results • While it imposes very few requirements on the DBMS, this solution suffers increases application development complexity

36. DB = Union(DB_Part_1, DB_Part_2, DB_Part_3, DB_Part_4) Shared Disk System DB_Part_3 DB_Part_1 DB_Part_2 DB_Part_4 Clusters for Scalability • Sybase will extend the continuous availability solution and the partition solution to provide cluster aware partition functionality • Data can be partitioned onto multiple nodes, but all partitions are accessible and available to all nodes via a shared disk system • Optimally, client applications are processed by the node that owns the partition required by the client application • However, in case of a failure, one or more of the remaining nodes seamlessly take over the partition(s) on the failed node to continue processing • Provides scalability and availability in a single solution

37. Security & Archiving • Encryption • Native Column-level Encryption in the Database • Auditing • Common Audit Repository • Enhancements to syb_audit functionality • Log Based or Sniffer based auditing • Archiving • Dynamic Archiving & Active Archiving

38. XML Engine Enhancements Multi Row (Multi Document) Index XML Schema support XML Data Type XQUERY Support for full XQUERY syntax User defined XQUERY functions XQUERY update syntax support XQUERY caches for efficiency Optimizer better tuned for XQUERY XML Transformation Enhancements SQLX scalar functions (XMLELEMENT etc.,) Cursor (row-by-row) mapping for SQL to SQLX SQLX mapping for SQL schemas and catalogs Tighter integration with Jallora Future XML Direction

39. ASE Real Time Services – Today & Future Power PocketBuilderApplication Development RepConnector ASE XML MobileAvantGo 3rd Party ASE Real Time Services Active Event Publisher Message Collector Message Bus (TIBCO JMS, RV, MQ)

40. Roadmap

41. ASE Roadmap Self Management & Data Mgmt Services Clusters & VLDB Low TCO • ASE Future • Clusters • Partitions • Real-time Analytics • Self Management (Phase 3) • Security – Native Encryption • ASE 12.5.x • Self Management (Phase 2) • Real Time Services • Federated Computing • Security – Auditing • XML – XQUERY • Dynamic Archiving • ASE 12.5.1 • Operational Scalability • Self Management • XML Performance & Transformations • Security – Authentication • Web Services 2004 2005 2006

42. Sybase ASEPerformance, Innovation at Lower Cost Most Successful Database Software Company in China -CCID Consulting ASE 15% lower TCO than Oracle - Standish Group “Sybase on Linux provides the stability, performance, and availability we need to keep our customers satisfied and our costs down.” - Iroquois Gas #1 in XML Storage Capabilities - ZapThink