32
Copyright © 2012, Oracle and/or its affiliates. All rights reserved. Insert Information Protection Policy Classification from Slide 13 1

Leveraging Oracle Database Partitioning and Hybrid Columnar Compression with Pillar Axiom QoS

Embed Size (px)

DESCRIPTION

Database partitioning was designed to enable old, stale data to be allocated to separate tablespaces to improve performance for more-recent “hot” data. So older data need not reside on fast, expensive media. The patented quality of service (QoS) of Oracle’s Pillar Axiom storage system enables partitioned data to both be moved to low-cost, high-capacity media and dynamically manage storage system resources (CPU, cache, I/O priority) so hot partitions receive a higher QoS (better performance) compared to “archive” partitions. Now add Oracle Database’s Hybrid Columnar Compression feature to the mix, and you have very space-efficient storage with an optimal allocation of resources to meet your business objectives at the lowest cost and highest performance. Have your cake and eat it too!

Citation preview

Page 1: Leveraging Oracle Database Partitioning and Hybrid Columnar Compression with Pillar Axiom QoS

Copyright © 2012, Oracle and/or its affiliates. All rights reserved. Insert Information Protection Policy Classification from Slide 131

Page 2: Leveraging Oracle Database Partitioning and Hybrid Columnar Compression with Pillar Axiom QoS

Copyright © 2012, Oracle and/or its affiliates. All rights reserved. Insert Information Protection Policy Classification from Slide 132

Leveraging Oracle Database Partitioning and Hybrid Columnar Compression with Pillar Axiom QoS

Dan MonksLead Architect, Axiom Storage North AmericaAaron NewcombSenior Principal Product Director, Storage

Page 3: Leveraging Oracle Database Partitioning and Hybrid Columnar Compression with Pillar Axiom QoS

Copyright © 2012, Oracle and/or its affiliates. All rights reserved. Insert Information Protection Policy Classification from Slide 133

Program Agenda

Data Growth and Complexity

Database Partitioning

Hybrid Columnar Compression

Axiom Quality Of Service

Page 4: Leveraging Oracle Database Partitioning and Hybrid Columnar Compression with Pillar Axiom QoS

Copyright © 2012, Oracle and/or its affiliates. All rights reserved. Insert Information Protection Policy Classification from Slide 134

Data Growth and Complexity

Average IT Database Growth 45% annually. Capital costs.

– Buy new hardware (storage, backup, infrastructure)

Operational costs.– Manually manage the data placement.

Performance implications of “more data”.– More cars on the highway means more traffic jams.

Page 5: Leveraging Oracle Database Partitioning and Hybrid Columnar Compression with Pillar Axiom QoS

Copyright © 2012, Oracle and/or its affiliates. All rights reserved. Insert Information Protection Policy Classification from Slide 135

2009

20122005 - 2010 2011

Q: What was the total sales revenue for Q3 in 2011?

Only the relevant partitions are accessed

Oracle Database Partitioning

Page 6: Leveraging Oracle Database Partitioning and Hybrid Columnar Compression with Pillar Axiom QoS

Copyright © 2012, Oracle and/or its affiliates. All rights reserved. Insert Information Protection Policy Classification from Slide 136

Oracle Database Partitioning

• Increases performance

• Increases availability

• Enables simpler management

• Enables storage cost optimization

• Transparent to applications Active

LessActive

Historical

Oracle Database Enterprise Edition

Partitioning Option

Page 7: Leveraging Oracle Database Partitioning and Hybrid Columnar Compression with Pillar Axiom QoS

Copyright © 2012, Oracle and/or its affiliates. All rights reserved. Insert Information Protection Policy Classification from Slide 137

ActiveLess

ActiveHistorical

Oracle Database Compression

Compress Large Application Tables

Compress All Data Types Compress Backup Data Typical Compression of 2-4X Transparent to applications

Oracle Database Enterprise Edition

Partitioning Option

AdvancedCompression

Compression4XUp to

Page 8: Leveraging Oracle Database Partitioning and Hybrid Columnar Compression with Pillar Axiom QoS

Copyright © 2012, Oracle and/or its affiliates. All rights reserved. Insert Information Protection Policy Classification from Slide 138

1 TBwith compression

10 TB of user dataRequires 10 TB of IO

100 GBwith partition pruning

20 GB with Storage Indexes

5 GB Smart Scan on Memory or Flash

Subsecond On Exadata Database

Machine

Less I/O = Better PerformanceUsing Partitioning, Compression Benefits Multiply

Page 9: Leveraging Oracle Database Partitioning and Hybrid Columnar Compression with Pillar Axiom QoS

Copyright © 2012, Oracle and/or its affiliates. All rights reserved. Insert Information Protection Policy Classification from Slide 139

ActiveLess

ActiveHistorical

Oracle Database Enterprise Edition

Partitioning Option

AdvancedCompression

HCC

Compression10XUp To

Data stored by columnand then compressed

Useful for data that is bulk loaded or moved

Query mode for data warehousing Typical 10X compression ratios Scans improve accordingly

Archival mode for old data Typical 15-50X compression ratios

• Transparent to applications Applications are unchanged

Choice of SATA or SAS drives

HyperActive

Compression50XUp To

Hybrid Columnar CompressionAvailable Only with Pillar Axiom or ZFS Storage Appliance

Page 10: Leveraging Oracle Database Partitioning and Hybrid Columnar Compression with Pillar Axiom QoS

Copyright © 2012, Oracle and/or its affiliates. All rights reserved. Insert Information Protection Policy Classification from Slide 1310

Oracle’s Hybrid Columnar CompressionUp to 50x Data Compression

Ideal for: Database & Data Warehouse Archiving, ILM, HA/DR

HCCCompressesDatabases

Exadata

Pillar Axiom ZFSSA

EMC and NetApp Require 3 to 5x more disk space 3 to 5x more power 3 to 5x more purchase price 3 to 5x more $ to run

Page 11: Leveraging Oracle Database Partitioning and Hybrid Columnar Compression with Pillar Axiom QoS

Copyright © 2012, Oracle and/or its affiliates. All rights reserved. Insert Information Protection Policy Classification from Slide 1311

Managing Database Data GrowthPartitioning With ACO + HCC

Before HCC, assuming a 50TB database:Hot Data (5%) = 2.5TBActive Data (20%) = 10TBHistorical Data (75%) = 37.5TBTotal DB Size = 50TB

After HCC, assuming a 50TB database:Hot Data (5%) = 2.5TBActive Data (20%) = 3TBHistorical Data (75%) = 1TBTotal DB Size = 6.5TB

Total reduction in Historical Data partition useable space needed is 97% !Total reduction in DB useable space needed is 87% !

2009

20122005 - 2010 2011

75% HCC 25% ACO

Page 12: Leveraging Oracle Database Partitioning and Hybrid Columnar Compression with Pillar Axiom QoS

Copyright © 2012, Oracle and/or its affiliates. All rights reserved. Insert Information Protection Policy Classification from Slide 1312

Oracle Pillar Axiom 600Quality of Service

Page 13: Leveraging Oracle Database Partitioning and Hybrid Columnar Compression with Pillar Axiom QoS

Copyright © 2012, Oracle and/or its affiliates. All rights reserved. Insert Information Protection Policy Classification from Slide 1313

Feature Function Benefits

Quality of Service Application prioritization and contention management

Business critical applications achieve the highest performance

Modular Architecture

Ability to dynamically scale both performance and capacity Investment protection

Partition and Hybrid Columnar

Compression

Storage and access data in the most efficient manner Highest performance and capacity

A Unique Axiom

Page 14: Leveraging Oracle Database Partitioning and Hybrid Columnar Compression with Pillar Axiom QoS

Copyright © 2012, Oracle and/or its affiliates. All rights reserved. Insert Information Protection Policy Classification from Slide 1314

Up to 4 ACTIVE-ACTIVE SLAMMERS with 64 Bricks

8 Control UnitsUp to 832 drivesUp to 1.6PB Capacity192GB Cache128 RAID ControllersSATA, FC, and SSD Storage Classes

SINGLE ACTIVE-ACTIVE SLAMMERwith ONE BRICK

2 Control Units13 Drives12TB Capacity48GB Cache

Pillar Axiom 600 features: Scalable System Architecture Patented Quality of Service (QoS) Software Engineered Integration with Oracle Software: Oracle Datagase, Oracle VM, Oracle Enterprise Manager, etc. Data Protection and Mobility: Copy Services (Vol Copy, Clones and Snapshots), SecureWORMfs, Replication Management: Application Profiles, Thin Provisioning, Storage Domains, MaxMan, Path Management Protocols: FC, iSCSI, CIFS, NFS

Oracle’s Pillar Axiom 600Enterprise SAN with Scalable Capacity and Performance

Page 15: Leveraging Oracle Database Partitioning and Hybrid Columnar Compression with Pillar Axiom QoS

Copyright © 2012, Oracle and/or its affiliates. All rights reserved. Insert Information Protection Policy Classification from Slide 1315

Virtual Server

Virtual Machine 2

Virtual Machine 1

Virtual Machine 3

FIFOQueue

Virtual Server

Virtual Machine 2

Virtual Machine 1

Virtual Machine 3

1 3

9

4 5

8

10

7

2 6

Premium Priority Queue

Medium Priority Queue

Low Priority Queue

Premium Medium Low Premium Medium Low

Typical Multi-Tier Array

4 5

8

10

1 3

7 9

2 6

Axiom ArchitecturePrioritizing the I/O Queue – Breaking the FIFO Model

Align The Business Value

Of The Application

To I/O Performance

Levels

Page 16: Leveraging Oracle Database Partitioning and Hybrid Columnar Compression with Pillar Axiom QoS

Copyright © 2012, Oracle and/or its affiliates. All rights reserved. Insert Information Protection Policy Classification from Slide 1316

Pillar Axiom: Patented QoS

QoS settings control drive type, count, and stripe placement

Deterministic IO Prioritization:The End of archaic FIFO queue management

Distributed RAID controllers add back-end bandwidth as capacity grows

Controls read/write bias and RAID type

Cache allocation implicitly controlled via QoS settings

Controls CPU cycles, IO priority, and RAID grouping

Page 17: Leveraging Oracle Database Partitioning and Hybrid Columnar Compression with Pillar Axiom QoS

Copyright © 2012, Oracle and/or its affiliates. All rights reserved. Insert Information Protection Policy Classification from Slide 1317 17

“Pillar Axiom 600 enables IT Administrators to utilize 80% of storage capacity without performance degradation – twice the industry average”

Best Storage Utilization with QoS

Source: Gartner Dataquest , 2010

Page 18: Leveraging Oracle Database Partitioning and Hybrid Columnar Compression with Pillar Axiom QoS

Copyright © 2012, Oracle and/or its affiliates. All rights reserved. Insert Information Protection Policy Classification from Slide 1318

Provisioning a Data RepositoryUsing an Application Profile

Data Type

StorageClass

LUN Performance Profile

Priority Access Bias I/O Bias

Control files SSD or FC High / Premium Mixed Mixed

Database Index SSD Medium Mixed Mixed

Database Tables SATA Medium Mixed Mixed

Temporary files SATA Medium Mixed Mixed

Online Redo Log Files FC High Sequential Write

Archive Log Files SATA Low Sequential Write

18

Admin needs more storage

Launches System Manager

Selects Application

Profile

Storage Is Provisioned

Page 19: Leveraging Oracle Database Partitioning and Hybrid Columnar Compression with Pillar Axiom QoS

Copyright © 2012, Oracle and/or its affiliates. All rights reserved. Insert Information Protection Policy Classification from Slide 1319

QoS MigrationsNo Interruption to Application I/O

Year End Activity

Weekend Lull

Reporting

Minimize Impact

Page 20: Leveraging Oracle Database Partitioning and Hybrid Columnar Compression with Pillar Axiom QoS

Copyright © 2012, Oracle and/or its affiliates. All rights reserved. Insert Information Protection Policy Classification from Slide 1320

Dom

ain 2

Dom

ain 1

Dom

ain 3

Create up to 64 Physical Domains in a Single Axiom• Refresh Legacy or Aging

Bricks without Disruption

• No Data Co-mingling: Isolate Application Data or Workloads to Physical Location

• Separate User Groups or Departments to Physical Location

• Separate Protocols (NAS or SAN) to Physical Drive Grouping Location

Extending QoS: Pillar Axiom Storage Domains

Page 21: Leveraging Oracle Database Partitioning and Hybrid Columnar Compression with Pillar Axiom QoS

Copyright © 2012, Oracle and/or its affiliates. All rights reserved. Insert Information Protection Policy Classification from Slide 1321

High VLUN

User Group 1 Domain User Group 3 Domain

Premium VLUN

Premium VLUN

Medium VLUN

Medium VLUN

Archive VLUN

Archive VLUN

High VLUN

User Group 2 Domain

SATA

SSD

SATA FC

SSD

Allows User Group application data to be isolated to specific bricks, thus eliminating any chance of inter-application I/O contention and provide charge-back capabilities.

User groups = Departments in an

internal cloud environment Customers in an external

cloud environment

Use Case: User Group Segregation

Page 22: Leveraging Oracle Database Partitioning and Hybrid Columnar Compression with Pillar Axiom QoS

Copyright © 2012, Oracle and/or its affiliates. All rights reserved. Insert Information Protection Policy Classification from Slide 1322

Partition Active and Archive

10 TB of user dataRequires 10 TB of IO

Apply ACO Compression4X Reduction

Apply HCC Compression20X Reduction

Data is 18x Smaller, Scans are 2000x faster

Putting It All TogetherUsing Partitioning, HCC, Axiom Benefits Multiply

ACO

HCC

High Priority

Medium Priority

Page 23: Leveraging Oracle Database Partitioning and Hybrid Columnar Compression with Pillar Axiom QoS

Copyright © 2012, Oracle and/or its affiliates. All rights reserved. Insert Information Protection Policy Classification from Slide 1323

Robert P Barracca JrDirector, Infrastructure TechnologiesAvero, LLC

Harnessing the Power of the Axiom 600 at Avero, LLC

Page 24: Leveraging Oracle Database Partitioning and Hybrid Columnar Compression with Pillar Axiom QoS

Copyright © 2012, Oracle and/or its affiliates. All rights reserved. Insert Information Protection Policy Classification from Slide 1324

Avero Company Overview

Avero helps restaurants sort through mountains of data to quickly find relevant, actionable insights to improve their business and outperform the competition

Avero is the leading provider of SaaS analytic applications and specialized services to the hospitality industry

Avero provides a solution comprised of software and services– Avero Slingshot business intelligence software

– Single Server Mentoring Performance Management solution

Avero processes over $14.6 billion in annual F&B revenues– Customers in over 32 countries on 5 continents

Get Relevant Insights Faster!

Page 25: Leveraging Oracle Database Partitioning and Hybrid Columnar Compression with Pillar Axiom QoS

Copyright © 2012, Oracle and/or its affiliates. All rights reserved. Insert Information Protection Policy Classification from Slide 1325

From the very beginning Avero has leveraged Oracle technology – Sun Servers running Solaris, Oracle Database, Oracle WebLogic, etc…

The addition of the Axiom storage array completed our implementation of Oracle across the stack

The Oracle Stack at Avero

Leveraging Oracle Technology

Page 26: Leveraging Oracle Database Partitioning and Hybrid Columnar Compression with Pillar Axiom QoS

Copyright © 2012, Oracle and/or its affiliates. All rights reserved. Insert Information Protection Policy Classification from Slide 1326

All of Avero’s products are built on Avero Data Warehouse

The Avero Data Warehouse contains over $180B in food and beverage revenue from almost 5,000 service outlets

Since 2005, average data growth has been 31% annually

Avero Data Warehouse

One Data Source

AveroData

Warehouse

Page 27: Leveraging Oracle Database Partitioning and Hybrid Columnar Compression with Pillar Axiom QoS

Copyright © 2012, Oracle and/or its affiliates. All rights reserved. Insert Information Protection Policy Classification from Slide 1327

Using ACO only would yield approximately 4x compression

Leveraging HCC allows for an average of 9x compression

– Archive Data – 26x

– Less Active Data – 10x

– Active Data – 8x

– Hot Data (ACO) – 4x

Managing Database Data Growth

Data Compression

Uncompressed ACO ACO+HCC

2000-2008 2009-2010 2011 2012

Page 28: Leveraging Oracle Database Partitioning and Hybrid Columnar Compression with Pillar Axiom QoS

Copyright © 2012, Oracle and/or its affiliates. All rights reserved. Insert Information Protection Policy Classification from Slide 1328

Primary database LUNs configured for highest priority

– FlashCache: Premium

– ASM DB tables: High

– ASM Online Recovery: Medium Production Test LUNs configured for

lower priority ZFS Mounts lowest priority

Leveraging Axiom QoS

Running Mixed Workloads

Page 29: Leveraging Oracle Database Partitioning and Hybrid Columnar Compression with Pillar Axiom QoS

Copyright © 2012, Oracle and/or its affiliates. All rights reserved. Insert Information Protection Policy Classification from Slide 1329

Learn More

www.oracle.com/storage– Whitepapers, Datasheets, Webinars

Videos– Youtube search for “Fact Finders” or “Oracle HCC”

Storage Training from Oracle University– education.oracle.com/storage

Page 30: Leveraging Oracle Database Partitioning and Hybrid Columnar Compression with Pillar Axiom QoS

Copyright © 2012, Oracle and/or its affiliates. All rights reserved. Insert Information Protection Policy Classification from Slide 1330

Like us on Facebook

Join the Conversation

• Engage with Oracle Experts

• Share knowledge and experiences

• Keep up with hardware topics & trends

• Access Oracle Hardware content easily, including events, webcasts, white papers and more

Join the Oracle Hardware Hub today

facebook.com/oraclehardware@aaronnewcomb

Page 31: Leveraging Oracle Database Partitioning and Hybrid Columnar Compression with Pillar Axiom QoS

Copyright © 2012, Oracle and/or its affiliates. All rights reserved. Insert Information Protection Policy Classification from Slide 1331

Page 32: Leveraging Oracle Database Partitioning and Hybrid Columnar Compression with Pillar Axiom QoS

Copyright © 2012, Oracle and/or its affiliates. All rights reserved. Insert Information Protection Policy Classification from Slide 1332