The New Storage Platform

PRESENTATION TITLE GOES HERE

.

.

Software Defined Storage

The New Storage Platform

Anil Vasudeva

President & Chief Analyst

IMEX Research

IT Industry Journey - Roadmap

Integrate Physical Infrast./Blades to meet CAPSIMS

Cost, Availability, Performance, Scalability, Inter-operability, Manageability & Security

Integration/Consolidation

Standard IT Infrastructure- Volume Economics HW/Syst SW

Standardization

Cloudization

On-Premises > Private Clouds > Public Clouds

DC to Cloud-Aware Infrast. & Apps. Cascade migration to SPs/Public Clouds.

Virtualization

Pools Resources. Provisions, Optimizes, Monitors

Automatically Maintains Application SLAs

(Self-Configuration, Self-Healing©IMEX_{, Self-Acctg. Charges etc.)}

Automation

IT Industry Roadmap

(Source: IMEX Research)

Big Data Analytics

Predictive Analytics - Unstructured Data

3

2014 Data Storage Innovation Conference. © 2010-14 IMEX Research. All Rights Reserved.

Enterprise VZ Data Center

On-Premise Cloud

Home Networks

Web 2.0

Social Ntwks.

Wireless

Internet

ISP

Core

Edge

ISP

ISP

ISP

ISP

ISP

Supplier/Partner

s

Remote/Branch Office

Public CloudCenter

ISP

Tier-3

Data Base

Servers

Tier-2 Apps

Management

Directory

Security

Policy

Middleware Platform

SCM, CRM Servers Database Servers, Middleware, Data Mgmt

Tier-1

Edge Apps

FC/IPSANs

ScaleOut NAS

R

A

I

D

S

O

F

T

W

A

R

E

D

I

S

K

C

O

N

T

R

O

L

L

E

R

H

O

S

T

High Availability

RAID-1

Cost

RAID-5

Performance

RAID-3

JBOD/DASD

RAID – First SW Defined Storage

1988

RAID SW Creates Specific Storage Capabilities (HA, Performance, Cost)

Sources: Vasudeva, Anil “A Case for Disk Arrays” Presented at LAN Conference, Santa Clara, CA Aug 1988

D

I

S

K

C

O

N

T

R

O

L

L

E

R

H

O

S

T

5

2014 Data Storage Innovation Conference. © 2010-14 IMEX Research. All Rights Reserved.

*IOPS for a required response time ( ms) *=(#Channels*Latency-1)

(RAID - 0, 3)

500

100

MB/sec

10

1

5

50

Data

Warehousing

OLAP

Business

Intelligence

IOPS*

(*Latency

-1)

Web 2.0

Audio

Video

Scientific Computing

Imaging

HPC

TP

HPC

10K

100 K

1K

100

10

1000 K

OLTP

eCommerce

Transaction

Processing

Workloads need Infrastructure > Optimized for Cost, Availability, Performance …

Virtualization led Workload Consolidation provided >20x Savings in RE, Pwr.,

Multiple VMs create I/O Blender Effect

Virtualized Server Penetration

0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 2011 2012 2013 2014 2015 2016

Virtualized Servers Penetration

WW Spending on Servers

995 Pre-Virtualization (VZ) Servers  78 VZ Servers

VZ SW & Support

Virtualization: TCO Savings

Power/Cooling & Data Management & Bottlenecks in Physical Servers sharing Memory

50% of all servers in 2011 growing to 85% by 2016

Virtualization: Impact on IT Infrastr.

7

2014 Data Storage Innovation Conference. © 2010-14 IMEX Research. All Rights Reserved.

NextGen Applications Exploding

0 50 100 150 200 250 2013 2014 2015 2016 2017

# Applications

(#K)

Applications Growth

2013-17

14 %

70 %

growth

Traditional Applications

NexGen Cloud

Applications

Traditional

Applications

NextGen Cloud

Apps

Application’s SLA dictates the Resources Required to meet specific

requirements of Availability, Performance, Cost, Security

, Manageability etc.

Platform Tools & Services / APIs

Operating Systems

Cloud Computing

Public Cloud Service

Providers

Private Cloud

Enterpris

e

App

_SLA

Applications/Software as a Service

PaaS

IaaS

SaaS

Virtualization

Resources (Servers, Storage, Networks)

App

_SLA

App

_SLA

App

_SLA

App

_SLA

Management

Hybrid

Cloud

9

2014 Data Storage Innovation Conference. © 2010-14 IMEX Research. All Rights Reserved.

9

Big Data Market Syst .& Services

Next IT Frontier: Big Data Analytics

Hadoop

NoSQL

RealTime DB

Graph DB

Next IT Frontier: Object Storage

11

2014 Data Storage Innovation Conference. © 2010-14 IMEX Research. All Rights Reserved.

11

11

HPC &

Commercial

Bioinformatics

& Healthcare

Entertainment-Social Netwks

Biz Intelligence

Big Data

Productivity/

_VDI

ITaaS: Apps leveraging SDDC/SDS

Virtualization Platform Has

•

Inherent knowledge of

Application’s Requirements

• Global View of Infrastructure

• Is Hardware agnostic

Converged Storage Pools

VM

App

OS

VM

App

OS

VM

App

OS

VM

App

OS

Virtualization Platform

vSphere / Hyper-V / Xen / KVM /…

13

2014 Data Storage Innovation Conference. © 2010-14 IMEX Research. All Rights Reserved.

Need: A New Storage Architecture

1 Key Tenets of Virtualization (VZ)

Resources Utilization

Costs

Utilization

63%

15%

Physical

Servers

Virtual

Servers

Compute

&

Networking

Storage Underperforms in VM Environments

• Very Random, Write intensive I/Os from some VMs

get blended with Sequential, Read Heavy I/Os from

other VMs resulting in:

o

Degraded Storage Performance by 30-50%

• Legacy Soln: Larger, more expensive storage

configs created to meet needed IOPs

Storage Capacity wastage

• Other Effects

o

Poor Thin Provisioning & Snapshots/Cloning

Inefficient VM Management

2 Storage Performance Issues in VZ

The VM I/O Blender – A key Culprit

3 Solution: Storage Defined Storage

Improves Perf., Mgmt, Cost/Provisioning,Snaps

• Create a storage abstraction layer

o

Do for Storage like Hypervisor for Compute

Virtualizes Storage for Optimum Mgmt.

• Unlock the Performance & Wasted Capacity of

Existing Storage by provisioning Storage as fast as

VMS can be created

o

Improves storage performance by 10x

Improves Thin Provisioning & Snapshots

Reduces capacity consumption up to 90%

• Provide a VM-Centric Management paradigm

o

VM-Centric Management

• Integrate Seamlessly into existing Hypervisor

Hypervisor

Hypervisor

Storage Hypervisor

VM 1

vDisk

VM 2

vDisk

VM 3

vDisk

VM n

vDisk

Storag

e Pool

Storage Hypervisor: Key to SDS

Storage Hypervisor Characteristics

Improves Perf., Mgmt, Cost/Provisioning,Snaps

• Software Solution that deploys in each Host

• Improves Storage Performance by up to 10x

• Reduces Capacity Cosumption by up to 90%

• Instantly provisions high performance Storage

• Provides VM-Centric Mgmt. for Ease-of-use

Hypervisor

Hypervisor

vDisk

vDisk

vDisk

VM n

VM 2

VM 1

Virtual

Storage

Appliance

Writes

vLogs

Reads

Optimized de-staging

vSpace

Storage Hypervisor Architecture

• Removes Random Write I/Os to remove storage

IO inefficiencies from Hypervisor

• Thin- Provisions vDisks to Outperform VMDKs

to cut storage costs in half

• Enable CDP making fast and scalable High

Performance Snapshots/Cloning

• Provides Instant Provisioning of High

Performance Storage

• Provides transparent VMDK level Mgmt.

• Lowers VDI Costs/desktop by 50%

15

2014 Data Storage Innovation Conference. © 2010-14 IMEX Research. All Rights Reserved.

Storage Hypervisor Overview

Storage

designed for

VMs

50% Lower

TCO

Dynamic,

Fast &

Resilient

•

Automated Storage Management

-Storage Policies

per VM derived

from

App/WorkLoad Requirements

•

Integrated with H

ypervisor &

Managed in vCenter

•

SLA

Requirements per VM managed

by Intelligent Data Placement

•

Instant

Storage Provisioning

•

Built-In Resiliency/Protection

from

HW failures

•

Dynamic Scalability

to Grow from

TB to PB

•

Automated Storage Management

-Storage Policies

per VM derived

from

App/WorkLoad Requirements

•

Integrated with H

ypervisor &

Managed in vCenter

•

SLA

Requirements per VM managed

by Intelligent Data Placement

•

Instant

Storage Provisioning

•

Built-In Resiliency/Protection

from

HW failures

•

Dynamic Scalability

to Grow from

Software Defined Storage:

17

2014 Data Storage Innovation Conference. © 2010-14 IMEX Research. All Rights Reserved.

VM Storage Policy

• Capacity

• Performance

• Availability

VM

VM

VM

VM

VM

VM

SDS Drives App-Centric Data Services

De-Dupe

Replication

Encryption

Snapshots

Compression

Caching

Back Up

Block Storage

File Storage

Object Storage

Control Plane

Data Plane

vSphere

vSAN/vVols

vCenter

Opns.Mgr.

Scale Out

NAS Storage

High End

Storage

Mid-Range

Storage

Open

Storage

1

2

1

2

vCloud Automation Center

19

2014 Data Storage Innovation Conference. © 2010-14 IMEX Research. All Rights Reserved.

VM

VM

VM

VM

VM

VM

VM

VM

VM

VM

VM

VM

VM

VM

VM

VM

VM

VM

VM

VM

VM

VM

21

2014 Data Storage Innovation Conference. © 2010-14 IMEX Research. All Rights Reserved.

VM Storage Policies Built in Advance

VM

VM Storage

Policy

_{• Capacity}

• Performance

• Availability

vCloud

Block Storage

File Storage

Object Storage

Control Plane

Data Plane

Scale Out

NAS Storage

High End

Storage

Mid-Range

Storage

OpenStack

Storage

REST APIs

_{REST APIs}

Infrastructure & Application Software Vendors

SDS is a Programmable Platform

23

2014 Data Storage Innovation Conference. © 2010-14 IMEX Research. All Rights Reserved.

vDataStore Capabilities visible to vCenter

1. At the time vSAN Cluster is

created, the capabilities

available in vSAN Datastores

are read and automatically

sent to vCenter

2. vCenter looks at these

capabilities available in

Datastores and sets VM

Storage Policies

3. VM Storage Policies then

have the Requirements of the

Application running on the VM

(Requirements such as

Availability, Performance and

Provisioning etc.)

New Role of IT: ITaaS

Virtual Workspace

Manages Access to Services, Applications and Data for Any Device

Virtual Workspace

Manages Access to Services, Applications and Data for Any Device

Hybrid Cloud

Seamlessly Extends Data Center to Public Cloud

Hybrid Cloud

Seamlessly Extends Data Center to Public Cloud

Software-Defined Data Center

Virtualizes the Entire Data Center

Automation & Management

Storage &

Availability

Compute

Network &

Security

25

2014 Data Storage Innovation Conference. © 2010-14 IMEX Research. All Rights Reserved.

ITaaS - From vCenter to vStores

VM

VM

VM

VM

VM

VM

VM

27

2014 Data Storage Innovation Conference. © 2010-14 IMEX Research. All Rights Reserved.

Outlook: Software Defined Storage

SDS - Key to “Next Big Thing in ITaaS”

Built-In strong features for Virtualized Data Centers & Cloud Federation

SDS Features

• Services Based Infrastructure provides Automation, Unifies

Control and Efficiency

• Provisioning via Policies and Workload-aware services to

match specific requirements of Each App

• Utilizes Open Standards and Interfaces based resilient,

commodity Data Storage deployable on any platform of choice

• Covers full spectrum of block, file and object storage

• Cost-effective and Highly Automated

PRESENTATION TITLE GOES HERE

.

Software Defined Storage

The New Storage Platform

Anil Vasudeva

President & Chief Analyst

IMEX Research