OpenStack Storage on Flash – Breaking Performance
and Economic BoundariesVenkat Kolli
Director Product Marketing
November 2015
SanDisk Confidential 2
Forward Looking Statements
This presentation contains certain forward-looking statements.
Any statement that refers to expectations, projections or other characterizations of future events or circumstances is a
forward-looking statement, including those relating to market growth for flash, industry trends including the increased
adoption of and demand for flash in various devices, enterprise applications and data centers, future memory
technology, technology transitions and future products, cost and timing of commercial availability. This presentation
also contains information from third parties, which reflect their projections as of the date of issuance of those
statements.
Actual results may differ materially from those expressed in these forward-looking statements due to factors detailed
under the caption "Risk Factors" and elsewhere in the documents we file from time to time with the SEC, including our
annual and quarterly reports.
We undertake no obligation to update these forward-looking statements, which speak only as of the date hereof or
the date of issuance by a third party, as the case may be.
SanDisk Confidential 3
Speaker
 Venkat Kolli
 Director of Product Management
 Enterprise Solutions
 Over 15 years of Storage Industry experience
 Introduced many innovative products at Fusion-io, NetApp, Veritas
 Developed storage solutions for Fortune 100 companies
 Venkat.kolli@sandisk.com
SanDisk Confidential 4
Agenda
 Flash introduction into data center
 OpenStack – New Cloud Operating System
 Storage for OpenStack
 SanDisk® all-flash storage system
 New Storage Architecture for New Worklaods – Benefits
 Performance advantage without premium costs
 Customer case study
SanDisk Confidential 5
A Global Fortune 500 Leader in Flash Storage Solutions
Personal Smartphones & Tablets Client SSDs Data Center SSDs, Systems & Software
Vertically Integrated
• Fab – Memory Cell Design
• Memory System Design
• Full Product Manufacturing
• Software Solutions
Patent Portfolio
• 5,000+ Patents
Financial Strength
Trailing 4 Quarter Financials*
• $6.1B Revenue
• $1.5B Net Cash
• $0.9B R&D Investment
SanDisk Ventures
• $75 Million for Strategic
Investments
Customers, Channels
• OEM
• Enterprise
• Hyperscalers
• Retail, Consumers
* Financials as of Q2’15. Net Cash = (Cash + cash equivalents + short-term and long-term marketable securities) less (debt at maturity value) as of end of Q2’15; patents are as of Q2 2015.
SanDisk Confidential 6
Datacenters are Increasingly Taxed to Provide Everything
with Minimal Budget
…with Quality of Service
Scale-out Capacity +
High Density
Performance
On-Demand
Low Operating Cost
(Low power, low cooling)
Simplicity
(Self-serve simple
administration)
SanDisk Confidential 7
Our All-Flash Designs Deliver Breakthrough Features and
Cost for an All-Flash Array
Low Power High Performance Scalable Reliable
Now
Breakthrough
Economics
The InfiniFlash™ System
8
 64-512 TB
JBOD of flash in 3U
 Up to 1M IOPS,
<1ms latency,
Up to 7 GB/s
throughput
 Energy Efficient
~400W power draw
 Connect
up to 8 servers
 Simple yet Scalable
SanDisk Confidential 9
From Server Virtualization to Cloud OS
SanDisk Confidential 10
OpenStack – New Cloud Operating System
11
Enabling Private Clouds with OpenStack
 OpenStack is the most predominant technology
used for Private Cloud build outs
 In less than 5 years ..
– 25,000 members
– 500 companies in 160 countries
– 20+ million lines of code
Why Private Cloud?
 Worlds largest brands on OpenStack
– Best Buy, Walmart
– AT&T, Comcast
– PayPal, Bloomberg, American Express, Fidelity
– Cisco Webex, Intel, TRUSTe, Cigna, Disney, …
SanDisk Confidential 12
Why OpenStack?
Ability to Innovate Less on Infrastructure maintenance and more on Applications and
Innovation
Open Technology Greater Flexibility, Interoperability, Faster Innovation
Cost Savings Eliminate most licensing costs, expensive support renewals.
ROI: 101% with 15.5 month payback2
Avoid Vendor Lock-in No Vendor lock-in for products, services, proprietary APIs and onerous
switching costs
1 Open Stack User Survey – Nov ‘14
2 A Forrester Total Economic Impact Study of OpenStack -July’ 15
Top Reasons for adopting OpenStack1
SanDisk Confidential 13
Ideal Storage for OpenStack
• Ceph – The most widely used OpenStack Block Storage
• 45% of OpenStack Block Storage runs on Ceph*
• More than just “another” OpenStack Driver
• Ceph is integral to OpenStack Cinder block storage
• Shares the same open source community development roots
• Overcome the performance limitations of OpenStack with
InfiniFlash HW optimizations
• Scale horizontally on demand- architecture built for scale
• Fully tested & proven for OpenStack deployments
• Hardened and tested for PB-scale deployments and
workloads
• Single vendor support for both hardware and software
+
SanDisk Confidential 14
One Storage Platform for All OpenStack Storage
Unified Scale-Out Storage System for Hyperscale
• Unified – One platform for all OpenStack workloads, iSCSI applications and more …
• Heterogeneous – InfiniFlash flexible architecture allows for performance tuned configurations for multiple workload
Cinder
Block Storage
Glance
Image Storage
Swift
Object Storage
Manila
File Storage
OpenStack Storage Interfaces
RBD Block Storage RGW Objects Ceph FS*
SanDisk Confidential 15
Disaggregation is the Key to Breakthrough Economics !
InfiniFlash™
Old Model
 Monolithic
 Proprietary storage OS
 Costly: $$$$$
New Model
 Disaggregated
 Open software stack
 Green !
 High performance
 Cost effective
Software Defined Storage
Standard X86 Servers
Advantages of
Disaggregation
 Lower TCA and TCO than
traditional models
 Rich choice of ecosystem
partners - compute,
networking and software
stacks
Evidence
 Decline of traditional
frame array business
SanDisk Confidential 16
InfiniFlash for OpenStack with Dis-Aggregation
 Compute & storage disaggregation
enables optimal resource utilization
- Allows for more CPU usage required for OSDs with
small block workloads
- Allows for higher bandwidth provisioning as required
for large object workload
 Independent scaling of compute and
storage
- Higher storage capacity needs do not force you to add
more compute and vice-versa
 Leads to optimal ROI for PB scale
OpenStack deployments
….
….
ComputeFarm
LU
N
LU
N
iSCSI Storage
…Obj Obj
Swift ObjectStore
…LU
N
LU
N
Nova with Cinder
& Glance
…
LibRBD
QEMU/KVM
RGW
WebServer
KRBD
iSCSI Target
StorageFarm
40GbE
network
SanDisk Confidential 17
Comprehensive Enterprise Storage Features
FILE SYSTEM *
BLOCK STORAGEOBJECT STORAGE
Keystone authentication
Geo-Replication
Erasure Coding
Striped Objects
Incremental backups
OpenStack integration
Configurable Striping
iSCSI
CIFS/NFS
Linux Kernel
Configurable Striping
S3 & Swift
Multi-tenant
RESTful Interface
Thin Provisioning
Copy-on Write Clones
Snapshots
Dynamic Rebalancing
Distributed Metadata
POSIX compliance
SanDisk focuses on flash optimization of Ceph and
lowering the cost of flash integration for Ceph * Future Ceph Roadmap
SanDisk Confidential 18
IF500 - Enhancing Ceph for Enterprise Consumption
SanDisk Ceph Distro adds usability and performance utilities
without sacrificing Open Source principles
All Ceph Performance improvements developed by SanDisk are contributed back to community
Open Source Ceph
with SanDisk
Performance
Improvements
 Out-of-the Box
configurations tuned for
performance with Flash
 Sizing & planning tool
 Higher node resiliency
with Multi-Path support
 Persistent reservations of
drives to nodes
 Ceph installer that is specifically built for InfiniFlash
 High performance iSCSI storage
 Better diagnostics with log collection tool
 Enterprise hardened QA @scale
 InfiniFlash drive management integrated into Ceph management (Coming Soon)
SanDisk Confidential 19
InfiniFlash Performance Advantage
900K Random Read Performance with 384TB of storage
Flash Performance unleashed
• Out-of-the Box configurations tuned for
performance with Flash
• Read & Write data-path changes for Flash
• x3-12 block performance improvement –
depending on workload
• Almost linear performance scale with
addition of InfiniFlash nodes
• Write performance WIP with NV-RAM
Journals• Measured with 3 InfiniFlash nodes with 128TB each
• Avg Latency with 4K Block is ~2ms, with 99.9 percentile
latency is under 10ms
• For Lower block size, performance is CPU bound at Storage
Node.
• Maximum Bandwidth of 12.2GB/s measured towards 64KB
blocks
S
SanDisk Confidential 20
InfiniFlash Performance Advantage
8K Read Performance improves up to 12x depending on the Block size
0
50000
100000
150000
200000
250000
300000
75:25 100:0
IOPS
% read : wrtie
8KB Random Block IOPS
IFOS 1.0
Stock Ceph
12x
IF500 Comparative Performance
Advantage
• Comparative performance of tuned IFOS
running on InfiniFlash vs. un-tuned stock
Ceph running on SSDs
• Ceph Giant release is used for this test
• Test on a single 512TB InfiniFlash node
SanDisk Confidential 21
InfiniFlash TCO Advantage
$-
$10,000,000
$20,000,000
$30,000,000
$40,000,000
$50,000,000
$60,000,000
$70,000,000
$80,000,000
Tradtional ObjStore on
HDD
IF500 ObjStore w/ 3
Full Replicas on Flash
IF500 w/ EC - All Flash IF500 - Flash Primary
& HDD Copies
3 year TCO comparison *
3 year Opex
TCA
0
20
40
60
80
100
Tradtional ObjStore on HDD IF500 ObjStore w/ 3 Full
Replicas on Flash
IF500 w/ EC - All Flash IF500 - Flash Primary & HDD
Copies
Total Rack
 Reduce the replica count with higher
reliability of flash
- 2 copies on InfiniFlash vs. 3 copies on
HDD
 InfiniFlash disaggregated architecture
reduces compute usage, thereby
reducing HW & SW costs
- Flash allows the use of erasure coded
storage pool without performance
limitations
- Protection equivalent of 2x storage with
only 1.2x storage
 Power, real estate, maintenance cost
savings over 5 year TCO
* TCO analysis based on a US customer’s OPEX & Cost data for a 100PB deployment
SanDisk Confidential 22
Before With IF500…..
 Customer deployed OpenStack-based private cloud
on Community Ceph
 Pain-points
• Cinder storage based on Ceph being used on
HDD+SSD could not get over 10K IOPS
• Limiting private cloud expansion from
including higher performance applications
 Solutions considered before IF500
• Alternate solutions with separate storage
architecture for high performance
applications add significant costs and higher
management overhead defeating the purpose
of OpenStack
 Meet higher performance goals with the same IF500
Ceph-based architecture without changing existing
infrastructure
• Ceph cluster on HDD will co-exist with IF500
Ceph cluster. Applications deployed on either
based on performance needs
• High performance workloads migrated to
InfiniFlash:
 Splunk log Analytics
 Apache Kafka messaging system
 Adding performance in lower TCO footprint
• 2 x 128TB IF500  expanding to 256TB next phase
• Expected to reduce real estate to < 1/3
• >50% power reduction expected
OpenStack Private Cloud on InfiniFlash IF500
Large Media Customer
SanDisk Confidential 23
The New Model
InfiniFlash™
 Disaggregate compute, storage
and networking leads to lower
costs
 All-Flash Storage performance
with costs of HDD solutions
 Open Source solutions with
Enterprise class reliability +
support
Standard X86 Servers
SanDisk Confidential 24
Faster time to insight and better business value
through disaggregated flash solutions
from SanDisk
SanDisk Confidential 25
© 2015-2016 Western Digital Corporation or its affiliates. All rights reserved. SanDisk and the SanDisk logo are trademarks of Western Digital Corporation or its affiliates, registered in the U.S. and other
countries. InfiniFlash is a trademark of Western Digital Corporation or its affiliates. Other brand names mentioned herein are for identification purposes only and may be the trademark(s) of their respective
holder(s).
@SanDiskDataCtr
@BigDataFlash itblog.sandisk.com
SanDisk Data Center Solutions

Deploying All-Flash Cloud Infrastructure without Breaking the Bank

  • 1.
    OpenStack Storage onFlash – Breaking Performance and Economic BoundariesVenkat Kolli Director Product Marketing November 2015
  • 2.
    SanDisk Confidential 2 ForwardLooking Statements This presentation contains certain forward-looking statements. Any statement that refers to expectations, projections or other characterizations of future events or circumstances is a forward-looking statement, including those relating to market growth for flash, industry trends including the increased adoption of and demand for flash in various devices, enterprise applications and data centers, future memory technology, technology transitions and future products, cost and timing of commercial availability. This presentation also contains information from third parties, which reflect their projections as of the date of issuance of those statements. Actual results may differ materially from those expressed in these forward-looking statements due to factors detailed under the caption "Risk Factors" and elsewhere in the documents we file from time to time with the SEC, including our annual and quarterly reports. We undertake no obligation to update these forward-looking statements, which speak only as of the date hereof or the date of issuance by a third party, as the case may be.
  • 3.
    SanDisk Confidential 3 Speaker Venkat Kolli  Director of Product Management  Enterprise Solutions  Over 15 years of Storage Industry experience  Introduced many innovative products at Fusion-io, NetApp, Veritas  Developed storage solutions for Fortune 100 companies  Venkat.kolli@sandisk.com
  • 4.
    SanDisk Confidential 4 Agenda Flash introduction into data center  OpenStack – New Cloud Operating System  Storage for OpenStack  SanDisk® all-flash storage system  New Storage Architecture for New Worklaods – Benefits  Performance advantage without premium costs  Customer case study
  • 5.
    SanDisk Confidential 5 AGlobal Fortune 500 Leader in Flash Storage Solutions Personal Smartphones & Tablets Client SSDs Data Center SSDs, Systems & Software Vertically Integrated • Fab – Memory Cell Design • Memory System Design • Full Product Manufacturing • Software Solutions Patent Portfolio • 5,000+ Patents Financial Strength Trailing 4 Quarter Financials* • $6.1B Revenue • $1.5B Net Cash • $0.9B R&D Investment SanDisk Ventures • $75 Million for Strategic Investments Customers, Channels • OEM • Enterprise • Hyperscalers • Retail, Consumers * Financials as of Q2’15. Net Cash = (Cash + cash equivalents + short-term and long-term marketable securities) less (debt at maturity value) as of end of Q2’15; patents are as of Q2 2015.
  • 6.
    SanDisk Confidential 6 Datacentersare Increasingly Taxed to Provide Everything with Minimal Budget …with Quality of Service Scale-out Capacity + High Density Performance On-Demand Low Operating Cost (Low power, low cooling) Simplicity (Self-serve simple administration)
  • 7.
    SanDisk Confidential 7 OurAll-Flash Designs Deliver Breakthrough Features and Cost for an All-Flash Array Low Power High Performance Scalable Reliable Now Breakthrough Economics
  • 8.
    The InfiniFlash™ System 8 64-512 TB JBOD of flash in 3U  Up to 1M IOPS, <1ms latency, Up to 7 GB/s throughput  Energy Efficient ~400W power draw  Connect up to 8 servers  Simple yet Scalable
  • 9.
    SanDisk Confidential 9 FromServer Virtualization to Cloud OS
  • 10.
    SanDisk Confidential 10 OpenStack– New Cloud Operating System
  • 11.
    11 Enabling Private Cloudswith OpenStack  OpenStack is the most predominant technology used for Private Cloud build outs  In less than 5 years .. – 25,000 members – 500 companies in 160 countries – 20+ million lines of code Why Private Cloud?  Worlds largest brands on OpenStack – Best Buy, Walmart – AT&T, Comcast – PayPal, Bloomberg, American Express, Fidelity – Cisco Webex, Intel, TRUSTe, Cigna, Disney, …
  • 12.
    SanDisk Confidential 12 WhyOpenStack? Ability to Innovate Less on Infrastructure maintenance and more on Applications and Innovation Open Technology Greater Flexibility, Interoperability, Faster Innovation Cost Savings Eliminate most licensing costs, expensive support renewals. ROI: 101% with 15.5 month payback2 Avoid Vendor Lock-in No Vendor lock-in for products, services, proprietary APIs and onerous switching costs 1 Open Stack User Survey – Nov ‘14 2 A Forrester Total Economic Impact Study of OpenStack -July’ 15 Top Reasons for adopting OpenStack1
  • 13.
    SanDisk Confidential 13 IdealStorage for OpenStack • Ceph – The most widely used OpenStack Block Storage • 45% of OpenStack Block Storage runs on Ceph* • More than just “another” OpenStack Driver • Ceph is integral to OpenStack Cinder block storage • Shares the same open source community development roots • Overcome the performance limitations of OpenStack with InfiniFlash HW optimizations • Scale horizontally on demand- architecture built for scale • Fully tested & proven for OpenStack deployments • Hardened and tested for PB-scale deployments and workloads • Single vendor support for both hardware and software +
  • 14.
    SanDisk Confidential 14 OneStorage Platform for All OpenStack Storage Unified Scale-Out Storage System for Hyperscale • Unified – One platform for all OpenStack workloads, iSCSI applications and more … • Heterogeneous – InfiniFlash flexible architecture allows for performance tuned configurations for multiple workload Cinder Block Storage Glance Image Storage Swift Object Storage Manila File Storage OpenStack Storage Interfaces RBD Block Storage RGW Objects Ceph FS*
  • 15.
    SanDisk Confidential 15 Disaggregationis the Key to Breakthrough Economics ! InfiniFlash™ Old Model  Monolithic  Proprietary storage OS  Costly: $$$$$ New Model  Disaggregated  Open software stack  Green !  High performance  Cost effective Software Defined Storage Standard X86 Servers Advantages of Disaggregation  Lower TCA and TCO than traditional models  Rich choice of ecosystem partners - compute, networking and software stacks Evidence  Decline of traditional frame array business
  • 16.
    SanDisk Confidential 16 InfiniFlashfor OpenStack with Dis-Aggregation  Compute & storage disaggregation enables optimal resource utilization - Allows for more CPU usage required for OSDs with small block workloads - Allows for higher bandwidth provisioning as required for large object workload  Independent scaling of compute and storage - Higher storage capacity needs do not force you to add more compute and vice-versa  Leads to optimal ROI for PB scale OpenStack deployments …. …. ComputeFarm LU N LU N iSCSI Storage …Obj Obj Swift ObjectStore …LU N LU N Nova with Cinder & Glance … LibRBD QEMU/KVM RGW WebServer KRBD iSCSI Target StorageFarm 40GbE network
  • 17.
    SanDisk Confidential 17 ComprehensiveEnterprise Storage Features FILE SYSTEM * BLOCK STORAGEOBJECT STORAGE Keystone authentication Geo-Replication Erasure Coding Striped Objects Incremental backups OpenStack integration Configurable Striping iSCSI CIFS/NFS Linux Kernel Configurable Striping S3 & Swift Multi-tenant RESTful Interface Thin Provisioning Copy-on Write Clones Snapshots Dynamic Rebalancing Distributed Metadata POSIX compliance SanDisk focuses on flash optimization of Ceph and lowering the cost of flash integration for Ceph * Future Ceph Roadmap
  • 18.
    SanDisk Confidential 18 IF500- Enhancing Ceph for Enterprise Consumption SanDisk Ceph Distro adds usability and performance utilities without sacrificing Open Source principles All Ceph Performance improvements developed by SanDisk are contributed back to community Open Source Ceph with SanDisk Performance Improvements  Out-of-the Box configurations tuned for performance with Flash  Sizing & planning tool  Higher node resiliency with Multi-Path support  Persistent reservations of drives to nodes  Ceph installer that is specifically built for InfiniFlash  High performance iSCSI storage  Better diagnostics with log collection tool  Enterprise hardened QA @scale  InfiniFlash drive management integrated into Ceph management (Coming Soon)
  • 19.
    SanDisk Confidential 19 InfiniFlashPerformance Advantage 900K Random Read Performance with 384TB of storage Flash Performance unleashed • Out-of-the Box configurations tuned for performance with Flash • Read & Write data-path changes for Flash • x3-12 block performance improvement – depending on workload • Almost linear performance scale with addition of InfiniFlash nodes • Write performance WIP with NV-RAM Journals• Measured with 3 InfiniFlash nodes with 128TB each • Avg Latency with 4K Block is ~2ms, with 99.9 percentile latency is under 10ms • For Lower block size, performance is CPU bound at Storage Node. • Maximum Bandwidth of 12.2GB/s measured towards 64KB blocks S
  • 20.
    SanDisk Confidential 20 InfiniFlashPerformance Advantage 8K Read Performance improves up to 12x depending on the Block size 0 50000 100000 150000 200000 250000 300000 75:25 100:0 IOPS % read : wrtie 8KB Random Block IOPS IFOS 1.0 Stock Ceph 12x IF500 Comparative Performance Advantage • Comparative performance of tuned IFOS running on InfiniFlash vs. un-tuned stock Ceph running on SSDs • Ceph Giant release is used for this test • Test on a single 512TB InfiniFlash node
  • 21.
    SanDisk Confidential 21 InfiniFlashTCO Advantage $- $10,000,000 $20,000,000 $30,000,000 $40,000,000 $50,000,000 $60,000,000 $70,000,000 $80,000,000 Tradtional ObjStore on HDD IF500 ObjStore w/ 3 Full Replicas on Flash IF500 w/ EC - All Flash IF500 - Flash Primary & HDD Copies 3 year TCO comparison * 3 year Opex TCA 0 20 40 60 80 100 Tradtional ObjStore on HDD IF500 ObjStore w/ 3 Full Replicas on Flash IF500 w/ EC - All Flash IF500 - Flash Primary & HDD Copies Total Rack  Reduce the replica count with higher reliability of flash - 2 copies on InfiniFlash vs. 3 copies on HDD  InfiniFlash disaggregated architecture reduces compute usage, thereby reducing HW & SW costs - Flash allows the use of erasure coded storage pool without performance limitations - Protection equivalent of 2x storage with only 1.2x storage  Power, real estate, maintenance cost savings over 5 year TCO * TCO analysis based on a US customer’s OPEX & Cost data for a 100PB deployment
  • 22.
    SanDisk Confidential 22 BeforeWith IF500…..  Customer deployed OpenStack-based private cloud on Community Ceph  Pain-points • Cinder storage based on Ceph being used on HDD+SSD could not get over 10K IOPS • Limiting private cloud expansion from including higher performance applications  Solutions considered before IF500 • Alternate solutions with separate storage architecture for high performance applications add significant costs and higher management overhead defeating the purpose of OpenStack  Meet higher performance goals with the same IF500 Ceph-based architecture without changing existing infrastructure • Ceph cluster on HDD will co-exist with IF500 Ceph cluster. Applications deployed on either based on performance needs • High performance workloads migrated to InfiniFlash:  Splunk log Analytics  Apache Kafka messaging system  Adding performance in lower TCO footprint • 2 x 128TB IF500  expanding to 256TB next phase • Expected to reduce real estate to < 1/3 • >50% power reduction expected OpenStack Private Cloud on InfiniFlash IF500 Large Media Customer
  • 23.
    SanDisk Confidential 23 TheNew Model InfiniFlash™  Disaggregate compute, storage and networking leads to lower costs  All-Flash Storage performance with costs of HDD solutions  Open Source solutions with Enterprise class reliability + support Standard X86 Servers
  • 24.
    SanDisk Confidential 24 Fastertime to insight and better business value through disaggregated flash solutions from SanDisk
  • 25.
    SanDisk Confidential 25 ©2015-2016 Western Digital Corporation or its affiliates. All rights reserved. SanDisk and the SanDisk logo are trademarks of Western Digital Corporation or its affiliates, registered in the U.S. and other countries. InfiniFlash is a trademark of Western Digital Corporation or its affiliates. Other brand names mentioned herein are for identification purposes only and may be the trademark(s) of their respective holder(s). @SanDiskDataCtr @BigDataFlash itblog.sandisk.com SanDisk Data Center Solutions