Trends in Business Continuity Planning RealTime Access to RealTime Information

1
Trends in Business Continuity PlanningReal-Time
Access to Real-Time Information

• Alok Pareek, VP Technology
• Feb 22, 2008

2
Agenda

• Introduction
• Trends in Business Continuity
• Availability, A 360 degree view
• Challenges, Approaches
• Data Freshness
• Real Time as a trend

3
Background

• Alok Pareek
• Vice-President Technology
• GoldenGate Software 3 years
• RD, Architecture, Real World Deployments
• Oracle Corp. Server Technologies Development
  10 years
• Recovery (Redo Generation and Write Ahead
  Logging)
• High Speed Data Movement (Cross Platform
  Transportable Tablespaces)
• Patent granted/filed key contributions to Oracle
  kernel include
• Multi-threaded redo generation (9i)
• Multiple block size cache support (9i)
• Cross Platform Transportable Tablespaces (Oracle
  10g)
• Whole database transport (10.2)
• Data Recovery/Repair (11g)
• Stanford University
• Research area - Recovery and Database Systems

4
About GoldenGate
Worldwide offices USA, EMEA, Asia Pacific, Latin
America
Exceptional customer support 24x7 global coverage
Company Strength and Service
Established in 1995
Rapid Growth in Strategic Partners
Established, Loyal Customer Base
5

Solutions

• HIGH AVAILABILITY
• REAL TIME DATA INTEGRATION

6
GoldenGate Customers Major Industries
Represented
Healthcare
Banking
Cable, Telco, and Manufacturing
Financial and Insurance Services
eBusiness, Retail, Public/Govt., Services
7
Business Continuity Trends a few points

• Broad Area
• Lots of papers, articles, books
• People/Process vs. Technology
• Trends are interesting
• Highlight Real Time Access to Real Time
  Information
• Link Business Continuity to Availability
• Link Availability to Access
• Link Access to Freshness
• Make the case for Real-Time
• Focus (today)
• Mission Critical Applications
• Transactional Data

8
A few Points to Consider

• Data produced by Humans vs. Machines
• Competition is healthy and encouraged,
  Globalization
• What does an 0800-1700 or a weekend mean anyway?
• Deletes/Updates are out of fashion, should they
  ever have been introduced?
• You want NOW but your pecking order is low

9
Just a decade ago
OLTP
OLAP/DW

• Mostly updates
• Many small transactions
• Mb-Tb of data
• Raw data
• Clerical users
• Up-to-date data
• Consistency, recoverability critical
• Gb - Tb of data

• Mostly reads
• Queries long, complex
• Gb - Tb of data
• Summarized, consolidated data
• Decision-makers, analysts as users
• Up-to-date data
• Consistency, recoverability critical
• Operational data

Hector Garcia Molina Stanford
Data Warehousing and OLAP
10
Where Real-Time Matters
Real-Time Access
Continuous Availability
Disaster Recovery
Disaster Tolerance
Continuous Operations
Tape Backups Disk Mirroring
Block-level Replication Hot Standby
Active-Active
Data Integration
BATCH Weeks / Days Custom Scripts
RIGHT TIME Hours ETL Scripts
REAL TIME Sub-Seconds TDM
NEAR REAL TIME Minutes / Seconds EAI
ETL Scripts
Sub-Seconds
Real-Time Information
Physical Infrastructure
Data
Transactional Operations
11
High Availability

• Definition
• Ratio of system uptime to sum of uptime and
  downtime
• Availability MTTF/(MTTFMTTR)
• Dependability
• Addressing Performance vs. Reliability in
  computer systems
• Hardware Faults, Software Bugs, Human errors are
  realities in any complex system deployment
• Enterprise applications need to function 24x7
• Disasters are no longer a distant threat
• Inadequate planning to handle outages
• Social Reputation, Competition, and the single
  click issue
• Direction
• Redefinition of Critical Systems, and uptime
  expectations
• From fault tolerance to reducing MTTR
• Batches to MicroBatches to continuous feeds

12
The 3 States of Availability 3600 View
A P P L I C A T IO N S
CRM, BILLING, SALES, TELCO, FIN
2 Planned Outage
3 Unplanned Outage
Unplanned outage
Data Failure
13
High Availability Concerns (No Outage)
1 Active

• Latency
• DSS vs. OLTP
• conflicting requirments
• Mixed workload
• Data validation
• Data Transformation

14
High Availability Concerns (Planned Outages)
2 Planned Outage
Unplanned outage
Common Approaches

• Selected windows of downtime
• Phased approach to maintenance

Migrations
Upgrades
Maintenance
15
High Availability Concerns (Unplanned Outages)

• Database Restore/Recovery
• RAID
• Shared Disk Clusters
• Standby database

3 Unplanned Outage
Common Approaches
16
Why are unplanned outages difficult? (1)

• Understanding failures
• Failure Points
• Statement
• Process
• Instance
• Database
• Site
• Failure Types
• Physical (Media, corruption, inconsistency
  amongst redundant copies)
• Logical (Incorrect DML, out-of-synch, accidental
  table drop)
• Failure Handling
• Automatic
• Manual

17
Why are unplanned outages difficult? (2)

• Mapping of symptoms to failure categories is
  complex
• Planning for all failure cases is non trivial
• Native repair solutions do not address complex or
  multiple failures
• Root cause analysis affects MTTR
• GOING FORWARD
• Failover, isolated repair will replace
  conventional recovery in computing environments
• Restores will be frowned upon
• System designers will increasingly focus on micro
  (surgical) repair

18
Ongoing HA Challenges

• Meeting Service levels due to performance impact
• Providing windows for Application, Database, OS
  upgrades/maintenance/migration
• Mapping of symptoms to failure categories is
  complex
• Planning for all failure cases is non trivial
• Native repair solutions do not address complex or
  multiple failures
• Root cause analysis affects MTTR
• GOING FORWARD
• Keeping OLTP really OLTP
• Rolling upgrade solutions will be expected of
  Application vendors
• Failover, isolated repair will replace
  conventional recovery in computing
  environments, Restores will be frowned upon
• System designers will increasingly focus on micro
  (surgical) repair

19
Keeping OLTP Systems Performing (State 1)

• Splitting workloads across multiple databases
  for
• Horizontal SCALABLITY for high throughput
  environments
• High speed, low-cost distributed caching
• Write/Read
• Multi Master load balancing solutions yielding
  better performance
• Avoid global serialization
• Faster response times (geographic proximity)
• Partitioning manageability advantages
• Offloading Reporting to real time copies
• Solution matters
• ROI on existing standby systems (hardware, or
  databases)
• BITS ? TRANSACTIONS ? APPLICATIONS?BUSINESS
  SOLUTION

20
Eliminating Planned Outage Windows (State 2)

• Addressing Upgrades/Migrations
• Rolling Database upgrades
• Zero-Downtime Platform migrations
• Zero-Downtime Database migrations
• Rolling Application upgrades
• Maintenance operations
• Index reorganizations
• Regenerating Statistics for Query optimizer
• Health Checks
• Verification, Validation

21
The rolling approach Technical challenges

• Data issues
• The secondary copy
• Instantiating Terabytes/Petabytes
• Staging areas
• Change Management
• Special Handling

• Synchronization issues
• Incremental data movement
• Transactional Integrity
• Reliable change data capture framework
• Source database impact
• Performance
• Application/Schema changes

• Failback strategy
• System/Application verification
• Continued data changes

22
Protection from Failures (State 3)

• Failover to synchronized target (Live)
• Multiple copies of data for high availability
  (temporal versions)
• Decouple Root Cause Analysis from MTTR
• Eliminate Restore from MTTR
• Application data already in cache increased
  response times
• Undo transactions using logical recovery features
• Recovery from user errors
• Increase availability by handling special class
  failures
• Redo block corruption
• Storage stack corruption
• Protection from lost writes from host

23
Availability and Data Freshness
To reduce latency and drive value, data
acquisition needs to approach real time.
Availability Requirements matter.
Business event
Data latency
Analysis latency
Business Value
Decision latency
Action Time
From TDWI The Business Case for Real-Time
BI Based on concept developed by Richard
Hackathorn, Bolder Technology
24
Summary

• Business Continuity from a technology standpoint
  is about addressing High Availability of
  Applications
• High Availability encompasses
• SLA on Production Systems
• Addressing Planned Outages without outage windows
• Addressing Unplanned Outages using Failover (all
  failures)
• Trends
• Continuous Application Availability
• Data Freshness across multiple systems
• Real Time Access to Real Time Information is the
  driver

25
Thank You.QAapareek_at_goldengate.com