Backup and Restore Practices for the Enterprise

STAN STRINGFELLOW is an independent contractor with 20 years of experience as a technical writer. He wrote the original software manuals for the Sun Enterprise 10000 (Starfire), and has done work for many different companies. He may be contacted at stan@stringfellow.com MIROSLAV KLIVANSKY manages the Storage Performance Engineering group at Sun Microsystems. He has researched I/O subsystem performance at Georgia Tech and has helped Sun customers plan Storage solutions for their data centers. He has published and taught in various performance-related areas, and has studied the performance of UNIX systems since 1993. MICHAEL BARTO is a technologist for Sun Microsystems Professional Services working as an IT architect with emphasis on Web-based collaboration and communication. He has over 15 years of system administration experience starting with his first Sun system in 1986.

1. Managing a Backup and Restore Initiative.


Sun and the Evolving Datacenter Model. Managerial Issues. Backup Tool Selection Criteria. Architectural Issues. Remote and Global Administration. Automation. High Availability. Performance. Ease-of-Use. Ease-of-Customization. Compatibility With Platforms and Protocols. Compatibility With Business Processes and Requirements. Backup Catalog Features. Tape and Library Support. Cost.



2. Backup and Restore Basics.


Physical and Logical Backups. Fully-Consistent File System Dumps. Full-Time Availability. Database Backup Technology. Online Database Backups. Offline Database Backups. Raw Device Backups. Advances in Backup Technology. Faster Throughput Rates. Greater Capacities. New Approaches to Online Backups With Database Technology. Automated Backup and Recovery Management Procedures. Multiplexing. Compression. Raw Backups With File-Level Restores. True Image Restore. Automatic Multistreaming.



3. Case Studies.


Case Study: A Transition to VERITAS NetBackup. Company Overview. Success, Growth, and Scalability Issues. Choosing the New Backup Tool. Hierarchical Architecture. Heterogeneous OS Support. Tape Library Support. API Available. Support for NAS and Thin Servers. Relationship with the Vendor. Complete Solution. Designing the New Architecture. Choosing the New Tape Libraries. Architecting Service Levels With the Business Units. Integrating Oracle into the Architecture. Scheduling Backup Jobs. Optimizing Recovery Time vs Backup Time. Architecting Tape Storage. Manual Backups vs Automatic Backups. Designing the Master/Slave Hierarchy. Integrating UNIX and NT Servers. VERITAS NetBackup FlashBackup. Multiplexing. Server-Independent Restores. Planning for SAN. Database Backups: They Like It Hot (sometimes). Two Backup Methods. Two Approaches to Backing Up Archive Logs. Database Backup Schedules. Tradeoffs Related to the New Architecture. Benefits Related to the New Architecture. Case Study: A Solstice Backup Implementation. Datacenter Overview. Worldwide Operations. Major Datacenter Applications. LeadGlobe Backup and Restore Architecture. Transition From a Mainframe Environment. Tape Libraries. Backup Servers. Storage Nodes and Private Subnets. SQL-BackTrack Fills Middleware Role. Backup and Restore Processes. Backup Schedules and Tape Rotations. Hot Backups, Cold Backups, and Mirrors. Vaulting Procedures. Multiplexing. Example Backup Architecture. Backup Components. Software. Hardware. Supported Servers. Support Applications. Example Backup Types and Backup Frequencies. Database Backups. Archive Log Backups. SQL-BackTrack Profile Backups. File System Backups. Secondary Backup Mechanisms. Example Backup Schedule. Example of Backup Architecture Map.



4. Methodology: Planning a Backup Architecture.


Introduction. Capacity Planning. Understanding the Environment. Dataset Size. Total Dataset Size. Size of Changes to the Dataset. Data Type. File Structure. Data Origin. Data Destination. The Data Path. Enterprise Backup Requirements. Expectations. Compression. Metadata Overhead. Recovery Performance. Ease of Use and Training Requirements. Simple Sizing Model. Workload Characteristics. Network Sizing. Channel Sizing. Disk Drive Sizing. Tape Drive Sizing. The Simple Sizing Model. Multiplexing as a Remedy. Network Remedies. Channel Remedies. Disk Remedies. Tape Remedies. Examples. Example 1: Local Attached Tape and Disk Drives. Example 2: Backup and Restore Across a LAN. Example 3: Backup and Restore Across a WAN. Other Considerations. Startup Time. CPU Sizing. Memory Sizing. UFS Direct I/O. Tape Back-Hitching. Tape Changing. Media Errors.



5. NetBackup Runbook.


Monitoring and Killing Backup Jobs. NetBackup Reports. Manual Backup and Restore. Interpreting Progress Logs.



6. Oracle Parallel Export Script Why a Parallel Export Script?


Overview of the paresh Approach. The list_tbls Script. The paresh Script.



Appendix A: High-Speed Database Backups on Sun Systems.


Executive Summary. Overview. Results. Conclusion. Technology for High Performance Backups. Hardware Configuration. Software Configuration. Delivering High-Speed Database Backups. A Range of Backup Solutions.



Appendix B: Business Continuity Planning and Sun Technologies.


Introduction. Planning for Continuity. Business Continuity Planning. Sun and BCP. Sun Enterprise Clusters Solutions. Dynamic System Domains. Sun StorEdgeÒ L700 Tape Library. Sun Data Center Consolidation Program. Continuity Alliances. Glossary.



Index.