WHITE PAPER

For many organizations, the purchase of additional storage hardware each year shows no sign of slowing. The user and depart-

mental need for more information and data appears insatiable,

so many IT organizations are forced to continually add more

storage. Regulatory requirements are also changing how and

why data is being retained, as organizations are required to

store more information for much longer periods. However,

this approach of constant storage spending is running up

against real-world budget restrictions, not to mention the

impact that data growth is having on response times and

application performance.

Research bears this fact out. In a recent survey of Oracle User

Group members, nearly 40% of the respondents indicate that

reducing costs is their primary reason behind implementing

tiered storage. But even with the right storage and compression

capabilities, deciding which data should reside where, and when

to migrate data from one tier to another, remains a serious chal-

lenge. That’s why forward-looking IT organizations are adopting

advanced strategies for managing data growth with innovative

database technologies that automate compression and storage

tiering while improving database performance.

Oracle Database 12c provides a rich feature set and database

options that optimize database storage to reduce continuous

demand and work together to help you implement an informa-

tion lifecycle management (ILM) solution in order to comply

with increasing data storage demands and new regulatory

requirements for data retention and protection.

This white paper provides details on the key features that are

used to optimize database storage for Oracle workloads and

can be used to implement an Oracle Database ILM solution.

ORACLE DATABASE 12C STORAGE OPTIMIZATION TECHNOLOGY Database storage optimization is a critical activity for nearly

every organization. The traditional approach of simply adding

more disk storage is becoming increasingly difficult to justify,

and cost is no longer the only issue. With Oracle Database

12c Storage Optimization, IT departments can manage the

impact of data growth with a unique set of technologies that

used together provide an effective approach to information

lifecycle management.

Improve Performance and Lower Costs with Oracle Database 12c Storage Optimization

COSTSPERFORMANCE

Data Partitioning Partitioning enables tiered storage, which allows data to

be saved to the most cost-effective storage for better

resource utilization.

n Oracle Partitioning option The use of Oracle Partitioning option enables large database

tables and indexes to be subdivided into smaller objects

that can be managed at a finer level of granularity to improve

data access performance.

Database Compression Technologies for all Data Types Oracle is a pioneer in database compression technology. More

than a decade ago, Oracle introduced Basic Table Compres-

sion, which compresses data at the block level, allowing you to

store up to 4x or more data per block using direct path opera-

tions. Today, Oracle Database 12c provides a comprehensive

set of compression capabilities across all types of data without

changes to your application to help customers improve query

performance while reducing storage costs.

n Advanced Row Compression Advanced Row Compression, a feature of Oracle Advanced

Compression, uses a unique compression algorithm specifi-

cally designed to work with OLTP applications that allows

for normal, conventional path operations to be compressed

without the limitations of direct path operations. It minimizes

the overhead of write operations on compressed data,

making it suitable for transactional and data warehouse

environments, extending the benefits of compression to

all application workloads. The benefits of Advanced Row

Compression go beyond disk storage cost savings. One

significant advantage is the ability to read compressed blocks

directly, in memory, without uncompressing the blocks.

This helps improve performance due to the reduction in I/O,

and the reduction in system calls related to the I/O opera-

tions. Further, the buffer cache becomes more efficient by

storing more data without having to add memory. Using the

Advanced Row Compression feature can provide a 2x to 4x

reduction in storage across all environments.

n Hybrid Columnar Compression Hybrid Columnar Compression is an approach Oracle devel-

oped for organizing data within a block. Unlike other orga-

nization methods that are either row or column based, this

new technology utilizes both row and column approaches for

organizing data. In this way, performance issues from using a

columnar organizational approach only is mitigated, while the

compression benefits of columnar storage are gained. With

IMPROVE PERFORMANCE AND LOWER COSTS WITH ORACLE DATABASE 12C STORAGE OPTIMIZATION 2

average storage savings of 10x, IT managers can dramatically

reduce and often eliminate the need to purchase additional

storage for several years. Hybrid Columnar Compression is

supported on the Oracle Exadata Database Machine, which

is designed for extreme performance. It is also supported on

Oracle Axiom and ZFS Storage Appliance (ZFSSA), helping to

further differentiate Oracle Engineered Systems.

Automatic Compression and Storage Tiering Oracle Advanced Compression incorporates a number of new

features that work together to optimize database storage. When

the use of the Heat Map feature, which collects data usage

information at the row and segment level, is combined with

the Automatic Data Optimization (ADO) feature, administrators

can create and apply ILM policies. This is based on the usage

patterns of their users’ data, and enables automatic compres-

sion and storage tiering. It also reduces storage costs and

improves application performance.

Heat Map Heat Map gives you a detailed view of how your data is being

accessed, and how access patterns are changing over time.

Automatic Data Optimization Automatic Data Optimization allows organizations to create poli-

cies for data compression and data movement and to implement

tiering of compression and storage. Oracle Database periodically

evaluates ADO policies, and uses the information collected by

Heat Map to determine which policies to execute.

SEEING IS BELIEVING!

Telling you how much you can save is one thing, but showing you your actual compression ratios, using your own data, is better. Oracle Database 12c users can take advantage of the Compression Advisor package that is bundled with the Oracle Database download to help make the business case for deploying Oracle Advanced Compression. This tool provides an estimate of the compression ratio and actual results that can be realized with Oracle Advanced Compression. The estimate is based on the analysis of a sample of data you provide to show the results you may obtain once you implement the Advanced Row Compression feature in your environment. See the Oracle Database documentation for more information about how to use the Compression Advisor.

Oracle Database Information Lifecycle Management Solution Oracle Database 12c Storage Optimization technologies address

the critical business problem of information lifecycle management.

The need for consistent management of information throughout

its useful life enables a business to utilize information more effec-

tively and helps meet compliance demands. Using a combination

of Advanced Data Optimization, Heat Map, Hybrid Columnar

Compression, and Advanced Compression and Partitioning

options, organizations can implement an automated, consistent,

and documented ILM approach for Oracle Database environments.

At the core of an Oracle Database ILM solution is the ability to

define multiple data classes and tiers of storage, and assign

different portions of data to different tiers based on the desired

cost, performance, and security for each portion. The key steps

in deploying an ILM solution using Oracle Database 12c include:n Define data classes or types of datan Create logical tiersn Define the data lifecycle

The first step to building an Oracle Database ILM solution is

to define the data classes or types of data that need lifecycle

management. To really understand your data you must analyze it

to determine which objects are associated with which applica-

tions, where those objects are physically located (what is the

class of storage), if the objects have been compressed, and the

granularity of the object (table vs. partitioned).

IMPROVE PERFORMANCE AND LOWER COSTS WITH ORACLE DATABASE 12C STORAGE OPTIMIZATION 3

Next, create logical storage tiers by identifying and creating

various storage tiers from expensive higher-performance storage

or high-capacity, lower-cost storage.

Finally, define the data lifecycle by describing how data migrates

across logical storage tiers over its lifetime. A lifecycle definition

may include one or more lifecycle stages that select a logical

storage tier based on data attributes such as compression, read-

only, or retention for data residing in that lifecycle stage.

PROVEN APPROACHES FOR ORACLE STORAGE OPTIMIZATION

There are a number of usage models that can result in improved

storage optimization using Oracle Database 12c. The most preva-

lent examples are included here.

By using Oracle Partitioning, more active data partitions can

be placed on faster, higher-performance storage where less

active and historic data can be placed on lower-cost storage.

Data compression can also be applied as desired on a partition-

by-partition basis—in both OLTP and data warehousing

applications. In this scenario, the business can meet all of its

performance, reliability, and security requirements, but at a

significantly lower cost than in a configuration where all data

is located on tier-one storage.

8 FIGURE 1: ORACLE DATABASE ILM SOLUTION

Figure 1 illustrates an Oracle Data-base ILM solution in which all the Oracle Database 12c Storage Optimi-zation features work together. In this example, Heat Map is tracking data access patterns, while ADO policies control compression and storage tier hierarchy. Table partitions are compressed with policies that are set for “compress for query” and “compress for archive,” with addi-tional policies for “tier to low cost” and “tier to archive.” Once set, ADO policies drive automatic compression and storage tiering as data access patterns are tracked in Heat Map, delivering an end-to-end Oracle ILM solution.

With OLTP applications, Advanced Row Compression is best

suited for the most active tables and partitions to ensure that

newly added or updated data will continue to be compressed as

data manipulation language (DML) operations are performed.

When using Oracle Exadata Database Machine or with Oracle

ZFS Storage Appliance or Axiom Storage, a combination of

Advanced Row Compression, Hybrid Columnar ”Warehouse”

Compression (also referred to as ”Query” Compression), or

Hybrid Columnar ”Archive” Compression can be utilized on a

partition-by-partition basis depending on how frequently the

data within the partition is accessed or manipulated.

Hybrid Columnar Warehouse Compression should be used with

tables and partitions that either aren’t being modified or are

being lightly modified. Hybrid Columnar Warehouse Compres-

sion is intended for query applications, but can tolerate data that

may be lightly updated using conventional DML operations such

as INSERT and UPDATE (with a reduced compression ratio on the

updated data). Hybrid Columnar “Archive” Compression is meant

for archive data that is read-only data and infrequently accessed.

The main purpose of “Archive” compression is to achieve very

high compression rates to reduce storage consumption.

Together these features provide a storage optimization solu-

tion that reduces storage costs, improves query performance

and automates the management of the data life cycle in Oracle

databases.

IMPROVE PERFORMANCE AND LOWER COSTS WITH ORACLE DATABASE 12C STORAGE OPTIMIZATION 4

For more information, visitwww.oracle.com/database12c

SOLUTION BENEFITS

SIMPLIFY ADMINISTRATION

One of the challenges with tiered storage architectures is the

need for subjective assumptions about how your data is used

rather than real knowledge. Heat Map changes that. ADO allows

you to use real usage information to create and enforce policies

about compression levels and storage tier movement. Once poli-

cies are in place and enforced, less administration is required and

focus can be on new projects for the business.

SAVE COSTSThe ability to reduce new storage expenditures for a poten-

tially multiyear period is a huge benefit. Many users of Oracle

Advanced Compression, for example, were able to reduce their

Oracle Database storage footprint by 50%. This adds up to signifi-

cant savings from mitigating new storage purchases, and poten-

tially decommissioning now unnecessary storage already in place

to save on maintenance contracts and other operating costs. In

addition, storage tiering can allow the organization to better align

storage costs with the performance needs of the data. Infre-

quently accessed data can be tiered to less costly storage and

expensive high performance storage can dedicated to the most

volatile and frequently accessed data.

IMPROVE PERFORMANCEOracle Database 12c Storage Optimization features can provide

tremendous performance benefits to a wide variety of applica tions

by improving manageability, performance, and availability. Oracle

Partitioning allows tables and indexes to be broken apart and

stored separately which greatly improves the flexibility of managing

very large tables and indexes. When combined with Oracle

Advanced Compression you can get even faster query perfor-

mance due to improved disk scan rates and I/O reductions. Oracle

Advanced Row Compression reduces the performance impact of

DML operations, plus Oracle operates directly on the compressed

format, which allows more data to be stored in memory and

reduces disk I/Os. For example, CSX Corporation improved batch

job completion time by 30% and improved its query performance

by 20%. This is clearly a benefit for the entire organization. At Land

O’Lakes, general ledger queries were completed 14x faster.

SUMMARYAs the growth of data and the storage infrastructure required to

host it continues unabated and at continually increasing costs,

organizations have found that developing a cost-effective approach

to database compression and storage tiering is essential.

Information lifecycle management solutions should enable

organizations to understand how their data is accessed overtime,

and manage data compression and storage tiering accordingly.

However, most ILM solutions for databases lack two key capa-

bilities: automatic classification of data, and automatic data

compression and movement across storage tiers.

Absent an ILM strategy, organizations face exploding storage

hardware costs and increasing complexity that will actually work

against the development of new IT solutions and the effective use

of those already deployed. Simply put, both the scale and growth

of data become overwhelming.

Oracle Database 12c provides an ideal platform for implementing

information lifecycle management.

Oracle Database 12c Storage Optimization features offer a

comprehensive set of ILM technologies that work together to

bring lasting value and true business benefit.