2012 Symantec Vision Everything You Need To Know About Oracle & Storage Foundation HA MITIGATE IT RISK WITH STORAGE FOUNDATION HA TO AVOID DOWNTIME AND ACCELERATE SERVICE RECOVERY Chris Atkinson Regional Product Manager - EMEA 1
2012 Symantec Vision Agenda IT Risk Management Strategies Solution Overviews Mitigating Risk 2 2
Understand SOURCES Of Risk & AGGREVATING Factors Human Error Change Processes Component Failure Environmental influence Poor Visibility Complexity Stress Location 60% of all unplanned downtime is caused by human error - IDC 95% of Environmental Disasters are Localized - Gartner Over 78% of companies do not test their DR plans - Gartner 3
ITRM Strategies Avoid Increase Infrastructure Visibility Adopt Common Tooling Standards Maintain Service Agility Intelligent Automated Operations 4
ITRM Strategies Mitigate MINIMISE Operational & Incident Impact Comprehensive Fault Detection Provide Redundancy Accelerate Full Service Recovery Provide PIT Recovery Mechanisms 5
ITRM Strategies Iterative CONTINUAL Process Improvement Pro-Actively Test & Analyse Organisational Learning Process Improvement 6
Overview Architecture Basics Oracle HA Architectures Oracle Fundamentals Symantec HA Architectures 7
Oracle HA Stacks 11gR2 Cache Fusion Low latency I/O & Fusion Flash I/O Cache Database Single Instance FAILOVER Failover: 5-20 Minutes + Simple App Design Simple Architecture Passive Licenses Saving Linear scale Database Parallel RAC Failover: 10 Seconds + Application Redesign Complex Architecture x2 Licence Cost Scale Overheads EXADATA Failover: 10 Seconds + RAC Configuration HCC ROI DW Specific Niche +XXX Cost Optimised storage 8
Understanding DB Recovery Instance Failure DOES result in an outage Redo Log MUST be applied Uncommitted transactions are rolled back Outage Time = Redo Log Delay CLIENT FAN Why Application Redesign For Recovery FAN Notifies status changes TAF/FCF Creates new connections Client MUST use Oracle Libraries FCF/TAF I/O Cache I/O Cache Shared Storage Redo Log Database Redo Log 9
ORACLE Symantec Solutions Veritas Cluster Server Integrated Storage Foundation Veritas Operation Manager Control & Visibility OpsCentre Oracle Enterprise Manager Database Management & Reporting 10
Symantec Solutions Cache Fusion Database Single Instance CFSHA Failover: 30+ seconds Simple App Design Simple Configuration Passive License Saving Linear scale Database Parallel SFRAC Failover: 10 Seconds + Encapsulates RAC Enhance Availability Enhance Performance Enhance Recovery Database CFSHA Violin Failover: <30 Seconds Extreme Performance Target Mixed/OLTP Faster Recovery Costly Storage Solution 11
Mitigate Reduce Incident Impact Provide Redundancy Comprehensive Fault Detection Accelerate Full Service Recovery Provide PIT Recovery Mechanisms 12
SAN Redundancy Impact Of Storage Node Outage Capacity degradation Entire service interrupted Uncommitted require resubmit Instance failure Application Outage/Service Capacity Lost (RAC) Storage Failure (e.g. LUN mapping, path fail) Redo log application ASM DB.dbf 13
SAN Redundancy Symantec I/O Shipping Technologies No Service interruption Minimises service capacity loss Interrupted writes captured Application Network I/O Shipping Allows processing to continue Storage Failure (e.g. LUN mapping, path fail) Avoids redo log replay DB.dbf Veritas Cluster File System 14 14
Mitigate Reduce Incident Impact Provide Redundancy Comprehensive Fault Detection Accelerate Full Service Recovery Provide PIT Recovery Mechanisms 15
Accelerate Recovery Service Recovery WEB APP DB Recovery Redo Instance Start Storage Recovery Detection DB REDO 16
Accelerate Instance Recovery DB Recovery Instance Recovery Delays Manual remediation often required Single Instance active ASM cost implication Redo log most significant Redo Instance Start SINGLE ANCE Storage Recovery DB REDO Dismount Deport Check Mount Import Detection Resource Coverage Resource Polling Instance Status? 17
Accelerate Instance Recovery Instant Recovery Process Start Comprehensive Fault Detection Greater Mitigation Chance Faster Resource Failover Instance Start RAC? Storage Recovery DB REDO Dismount Deport Check Mount Import Detection All Resources Disk Intelligent Monitoring Framework Deep Monitoring SQL Instance 18
Accelerate Instance Recovery Extreme I/O Benefits Recovery! Faster Failover Higher Performance AND Consolidation DB Recovery Redo Smaller I/O Cache I/O Cache Smaller Redo Log Reduced Delay Cluster File System HA Database Redo Log 19
Accelerate Service Recovery Full Service Recovery Challenges High Operational Risk & Demand Significant Service Recovery Delays Service Recovery Service Recovery DR Recovery Lacks Service Control Fault Location? WEB APP Service Prioritisation Infrastructure Switch? DB Handoff Delays Scripted Processes 20
Accelerate Service Recovery Automate HA&DR Service Recovery Recover Full Services Faster Avoid Operational Risk Reduce Complexity Risk Service Recovery Veritas Operations Manager (VBS) Veritas Operations Manager (DR Plan) Multi-tier Control WEB Wizard Process VBS Validation APP Automated Recovery Component Status DB Control External Elements Prioritise VBS 21
High Availability Choices Customer Decision Criteria Recovery Time Objective? Workload Capacity High Medium Criticality & Performance Requirements Low RAC Savings Single Instance Symantec & Violin High Availability Solutions 22
Then Wrap Up... 23
Symantec Solutions Cache Fusion Database Single Instance CFSHA Failover: 30+ seconds Simple App Design Simple Configuration Passive Licenses Saving Linear scale Database Parallel SFRAC Failover: 10+ Seconds Encapsulates RAC Enhance Availability Enhance Performance Enhance Recovery Database CFSHA Violin Failover: <30 Seconds Extreme Performance Target Mixed/OLTP Faster Recovery Costly Storage Solution 24
Mitigate MINIMISE Operational & Incident Impact Avoid Downtime Faster Service Recovery 25
2012 Symantec Vision Thank You! START AVOIDING DOWNTIME AND ACCELERATE SERVICE RECOVERY TODAY! www.symantec.com/database-management Chris Atkinson Chris_atkinson@symantec.com 26