The Challenges of Managing and
Protecting Expodential Data Growth
Robert Amatruda, Research Director Data
Protection and Recovery
The Challenges of Managing and
Protecting Expodential Data Growth
 Customers increasingly managing not just information but
infrastructure
 Increasingly, geo-dispersed data drive the need for flexible
and agile infrastructure
 Rapid unstructured data growth as a result of
mobility, social media, big data and cloud adoption
 Data growth, virtualization, optimization and edge protection
coupled with more aggressive RTO/RPO’s driving more use
of disk and replication

Š IDC Visit us at IDC.com and follow us on Twitter: @IDC

2
Explosion of Devices and Management
Challenges in the Datacenter
Data centers are the cornerstone of business…

…managing both infrastructure and information.
The Evolving Data Landscape
App
Data

User
Data

Ever-growing data sets require
dispersed data management
 Workloads need to run where the
data is located
 Geo-dispersed data sets require
geo-dispersed workloads

Mobile
Data
Decentralized
data stores

 Workloads are gaining location
awareness
 Social media, big data and cloud
usage is generating more
unstructured data
 Machines generating more semistructured

IDC Confidential

4
Unstructured Data is Growing Rapidly
 Implications of Unstructured Data
growth:
 Drives the need for capacity
optimized systems that house
unstructured data

 Unstructured data can be
stored/accessed via onpremises or cloud-based
delivery models
 Object-based storage platforms
also power the cloud storage-as-aservice solutions

140

120
100
80

60
40
20

0
2009

2010

2011

2012

2013

2014

2015

2016

2017

Capacity Optimized (shipped PB)
Performance Optimized (shipped PB)

• UNSTRUCTURED DATA accounts for
70-80% of storage capacity growth
• Capacity-optimized storage
spending growth has 16.2% CAGR

5
Top Unstructured Data Use-Cases
How is your organizations’ file-based data split amongst these following types of
files?
File and print services or network
shares

31%

Virtual server or desktop images

14%

Back up/Recovery

19%

CAD/Images (pictures, X-rays,
records)

8%

Video/audio (surveillance, training,
entertainment)

7%

Archive

16%

Big Data (Hadoop Clusters)

5%

Others

1%
0%

10%
2013

20%
% of respondents
2012

30%

40%

N=150

Š IDC Visit us at IDC.com and follow us on Twitter: @IDC

6
Changes in the Protection and Recovery
Landscape
Today …

In the Past …








Backup
Back-office
Structured
Datacenter
Data only
100’s of GB
Initial SANs

Performance









In the Future …

+ Options
C-level focus
+ Unstructured
+ Edge
Data and images
Multi-TBs
Multi-approach









Physical
Virtual
Consolidation
 What’s driven these changes?

Cloud
as-a-Service
++ Objects
++ Endpoints
Convergence
Petabytes
Reuse freedom

Unified Mgmt
7
Factors that have Driven Transformation

Backup

Replica

DR

1. Virtualization  Images and Data
2. Cloud Era  Storage to/from the Cloud
vBackup

Snaps

Archive

3. Data Growth  B2D, Storage Efficiency
4. Recovery  Snaps, Replicated Backup
5. Better Retrieval  Items, indexing

Disk as a
Transformative
Technology

8
Factors that will Drive Transformation

1. Control v. Data Movement 
Technology Integration

Unified Data
Management

2. Copy Data reuse  Native format

Backup

Replica

DR

3. Role of Flash/SSD  Performance

Search

Snaps

Archive

4. Scale out Storage Architectures 
Consolidation
5. Image mobility  Recovery to/in the
Cloud

9
Questions

Robert Amatruda
ramatruda@idc.com
10
Spectra Logic

Next Generation Technologies
What did we just hear from IDC?
• Unstructured data drives 70-80% of storage growth
– Unstructured Data = Files… lots of them!

• What’s driving it?
– Virtualization
– Backup/Recovery/Archive
– Big Data
– Video/audio

• And this trend will continue into the future
nTier Verde –
Simply Affordable File Storage
• No previous storage experience required
• Half the cost of traditional file storage
• Never lose data
No previous storage experience required
• From power on to production in 10 minutes
– Simple user interface
– Minimal formatting of disks required

• Monitoring/reporting
– Visual Status Beacon
– Automatic alerts
– Call home
Half the cost of traditional file storage
• Enterprise capacity storage for the cost of JBOD
• Install it yourself
• PriceLock for support pricing
• Compression
Subsystem Category

$ / GB

Solid State Disk (DRAM)

$400 - $1000

Flash

< $50

Enterprise Disk Subsystem

$5 - $16

Midrange Disk Subsystem

$1 - $6

Economy Disk Subsystem

< $.50

Automated Tape Library

$.05 - $1
Source: Horison, Inc.
Never lose data
• Hardware
– Enterprise-class, capacity drives
– Burn-in at our factory to eliminate “infant mortality” of
components
– Redundant power/fans

• Software
– Triple parity option for RAID
– No RAID write hole issue
– Intelligent rebuilds
nTier Verde
nTier Verde 2U
Supported Hard Drives

nTier Verde 4U

4 TB 7200 RPM SAS Hard Drives

Master Node Capacity*

6 to 11 Drives
24 TB to 44 TB RAW

10 to 35 Drives
40 TB to 140 TB RAW

Expansion Node Support

1 Expansion Node

9 Expansion Nodes

Max Capacity*

220 TB RAW Max

1.7 PB RAW Max

Parity Options

Mirroring, Single Parity, Double Parity, Triple Parity
Double Parity recommended for most use cases

Hot Spares

Unlimited

Supported Protocols

NFS, CIFS

3 x 1 GigE Data Ports

Standard

Standard

2 x 10 GigE Data Ports

Optional

Standard
* Using 4 TB drives
Store massive data forever at virtually no cost

• Absolute lowest cost way to store data
• Easiest way to move bulk data
• Most scalable long-term storage available
Absolute Lowest Cost Way to Store Data

Spectra BlackPearl plus
T950 with TS1140 drives

Spectra BlackPearl plus
T950 with LTO drives

Spectra BlackPearl plus
T380 with LTO drives

6.4 PB uncompressed

2.4 PB uncompressed

1.9 PB uncompressed
Easiest Way to Move Bulk Data
Spectra’s DS3 interface makes it possible
• Industry first interface to tape via a web browser
• RESTful (the way the web works)
• Modeled after Amazon S3

FC
SAS

10 GigE
Client
Software

Standard

Object

Interface

Inside BlackPearl:
 SSD
 Object tracking
 Tape Library
Management
 LTFS
 Data Integrity
Verification
 Data Security
 Integration
with BlueScale

Objects
to
Deep
Storage
Most scalable long-term storage available

• LTFS for standard reading/writing
• Scale from TB to Exabytes
• Keep data forever w/ media migration
Questions?

Eliminating the Problems of Exponential Data Growth, Forever

  • 1.
    The Challenges ofManaging and Protecting Expodential Data Growth Robert Amatruda, Research Director Data Protection and Recovery
  • 2.
    The Challenges ofManaging and Protecting Expodential Data Growth  Customers increasingly managing not just information but infrastructure  Increasingly, geo-dispersed data drive the need for flexible and agile infrastructure  Rapid unstructured data growth as a result of mobility, social media, big data and cloud adoption  Data growth, virtualization, optimization and edge protection coupled with more aggressive RTO/RPO’s driving more use of disk and replication © IDC Visit us at IDC.com and follow us on Twitter: @IDC 2
  • 3.
    Explosion of Devicesand Management Challenges in the Datacenter Data centers are the cornerstone of business… …managing both infrastructure and information.
  • 4.
    The Evolving DataLandscape App Data User Data Ever-growing data sets require dispersed data management  Workloads need to run where the data is located  Geo-dispersed data sets require geo-dispersed workloads Mobile Data Decentralized data stores  Workloads are gaining location awareness  Social media, big data and cloud usage is generating more unstructured data  Machines generating more semistructured IDC Confidential 4
  • 5.
    Unstructured Data isGrowing Rapidly  Implications of Unstructured Data growth:  Drives the need for capacity optimized systems that house unstructured data  Unstructured data can be stored/accessed via onpremises or cloud-based delivery models  Object-based storage platforms also power the cloud storage-as-aservice solutions 140 120 100 80 60 40 20 0 2009 2010 2011 2012 2013 2014 2015 2016 2017 Capacity Optimized (shipped PB) Performance Optimized (shipped PB) • UNSTRUCTURED DATA accounts for 70-80% of storage capacity growth • Capacity-optimized storage spending growth has 16.2% CAGR 5
  • 6.
    Top Unstructured DataUse-Cases How is your organizations’ file-based data split amongst these following types of files? File and print services or network shares 31% Virtual server or desktop images 14% Back up/Recovery 19% CAD/Images (pictures, X-rays, records) 8% Video/audio (surveillance, training, entertainment) 7% Archive 16% Big Data (Hadoop Clusters) 5% Others 1% 0% 10% 2013 20% % of respondents 2012 30% 40% N=150 © IDC Visit us at IDC.com and follow us on Twitter: @IDC 6
  • 7.
    Changes in theProtection and Recovery Landscape Today … In the Past …        Backup Back-office Structured Datacenter Data only 100’s of GB Initial SANs Performance        In the Future … + Options C-level focus + Unstructured + Edge Data and images Multi-TBs Multi-approach        Physical Virtual Consolidation  What’s driven these changes? Cloud as-a-Service ++ Objects ++ Endpoints Convergence Petabytes Reuse freedom Unified Mgmt 7
  • 8.
    Factors that haveDriven Transformation Backup Replica DR 1. Virtualization  Images and Data 2. Cloud Era  Storage to/from the Cloud vBackup Snaps Archive 3. Data Growth  B2D, Storage Efficiency 4. Recovery  Snaps, Replicated Backup 5. Better Retrieval  Items, indexing Disk as a Transformative Technology 8
  • 9.
    Factors that willDrive Transformation 1. Control v. Data Movement  Technology Integration Unified Data Management 2. Copy Data reuse  Native format Backup Replica DR 3. Role of Flash/SSD  Performance Search Snaps Archive 4. Scale out Storage Architectures  Consolidation 5. Image mobility  Recovery to/in the Cloud 9
  • 10.
  • 11.
  • 12.
    What did wejust hear from IDC? • Unstructured data drives 70-80% of storage growth – Unstructured Data = Files… lots of them! • What’s driving it? – Virtualization – Backup/Recovery/Archive – Big Data – Video/audio • And this trend will continue into the future
  • 13.
    nTier Verde – SimplyAffordable File Storage • No previous storage experience required • Half the cost of traditional file storage • Never lose data
  • 14.
    No previous storageexperience required • From power on to production in 10 minutes – Simple user interface – Minimal formatting of disks required • Monitoring/reporting – Visual Status Beacon – Automatic alerts – Call home
  • 15.
    Half the costof traditional file storage • Enterprise capacity storage for the cost of JBOD • Install it yourself • PriceLock for support pricing • Compression Subsystem Category $ / GB Solid State Disk (DRAM) $400 - $1000 Flash < $50 Enterprise Disk Subsystem $5 - $16 Midrange Disk Subsystem $1 - $6 Economy Disk Subsystem < $.50 Automated Tape Library $.05 - $1 Source: Horison, Inc.
  • 16.
    Never lose data •Hardware – Enterprise-class, capacity drives – Burn-in at our factory to eliminate “infant mortality” of components – Redundant power/fans • Software – Triple parity option for RAID – No RAID write hole issue – Intelligent rebuilds
  • 17.
    nTier Verde nTier Verde2U Supported Hard Drives nTier Verde 4U 4 TB 7200 RPM SAS Hard Drives Master Node Capacity* 6 to 11 Drives 24 TB to 44 TB RAW 10 to 35 Drives 40 TB to 140 TB RAW Expansion Node Support 1 Expansion Node 9 Expansion Nodes Max Capacity* 220 TB RAW Max 1.7 PB RAW Max Parity Options Mirroring, Single Parity, Double Parity, Triple Parity Double Parity recommended for most use cases Hot Spares Unlimited Supported Protocols NFS, CIFS 3 x 1 GigE Data Ports Standard Standard 2 x 10 GigE Data Ports Optional Standard * Using 4 TB drives
  • 18.
    Store massive dataforever at virtually no cost • Absolute lowest cost way to store data • Easiest way to move bulk data • Most scalable long-term storage available
  • 19.
    Absolute Lowest CostWay to Store Data Spectra BlackPearl plus T950 with TS1140 drives Spectra BlackPearl plus T950 with LTO drives Spectra BlackPearl plus T380 with LTO drives 6.4 PB uncompressed 2.4 PB uncompressed 1.9 PB uncompressed
  • 20.
    Easiest Way toMove Bulk Data Spectra’s DS3 interface makes it possible • Industry first interface to tape via a web browser • RESTful (the way the web works) • Modeled after Amazon S3 FC SAS 10 GigE Client Software Standard Object Interface Inside BlackPearl:  SSD  Object tracking  Tape Library Management  LTFS  Data Integrity Verification  Data Security  Integration with BlueScale Objects to Deep Storage
  • 21.
    Most scalable long-termstorage available • LTFS for standard reading/writing • Scale from TB to Exabytes • Keep data forever w/ media migration
  • 22.