SlideShare a Scribd company logo
DataCore™Virtual SAN introduces the next evolution in Software-defined
Storage (SDS) by creating high-performance and highly-available shared
storage pools using the disks and flash storage in your servers. It addresses
the requirements for fast and reliable access to storage across a cluster of
servers at remote sites as well as in high-performance applications.
DataCoreVirtual SAN virtualizes the local storage on two or more physical x86-
64 servers. It can leverage any combination of magnetic disks (SAS, SATA) and
optionally flash, to provide persistent storage services as close to the application
as possible without having to go out over the wire (network or fabric).Virtual disks
provisioned from DataCoreVirtual SAN can also be shared across the cluster to
support the dynamic migration and failover of applications between hosts.
DataCoreVirtual SAN addresses the challenges that exist today within
many IT organizations such as single points of failure, poor application
performance (particularly within virtualized environments), low storage
efficiency and utilization, and high infrastructure costs.
Let’s explore how DataCoreVirtual SAN addresses each of these challenges
with a proven, 10th generation software-defined storage solution, purpose-
built for maximizing storage resources across the enterprise.
Virtual SAN - A Deep Dive
into Converged Storage
DataCoreTechnical Whitepaper
Virtual SAN Benefits
Only 2 servers needed
for High-Availability
RAM for I/OAcceleration
Flash is Optional
100M IOPS, 64PB with
64 Host Cluster
Works with all major
Hypervisors
Integrate Converged,
SAN, and Cloud Storage
WHITEPAPER
2
Synchronous Mirroring
The availability of business-critical data is extremely important
for your organization. DataCoreVirtual SAN provides protection
against catastrophic events; outages that fault-tolerance
measures alone cannot adequately protect against. Synchronous
mirroring provides data-level redundancy to maximize data
availability and keep applications running without interruption.
DataCoreVirtual SAN accomplishes high-availability within the
cluster by synchronously mirroring every data block between
two fully active cluster nodes, where others may require three or
more hosts to accomplish that same capability.The two nodes
maintain redundant copies of the data ensuring that even if
one cluster node becomes inoperable, the applications will not
experience any interruption, and the data is always available.
Synchronous mirroring also plays a role in providing redundant
write cache.When running in a mirrored configuration, a high-
speed write cache is enabled.Writes are only confirmed when
both nodes acknowledge they have received the data. If a partner
node becomes unavailable due to maintenance for example,
then write data is still safe and destaged to persistent storage.
Unlike other solutions, where performance degrades significantly
when a loss of redundancy occurs, DataCoreVirtual SAN read
cache and write-coalescing functions are still enabled, providing
read acceleration and optimized destaging operations.
Below is a diagram that outlines the order of operations
that occurs during synchronous mirroring:
WHITEPAPER
1.	 Inbound write I/O received
from Host at Site A
2.	 Inbound write I/O committed to
high-speed RAM cache on SSV-1
3.	 Inbound write I/O transferred to SSV-2
4.	 Inbound write I/O committed to
high-speed RAM cache on SSV-2
5.	 Acknowledgement of write I/O
receipt from SSV-2 to SSV-1
6.	 Acknowledgement of write I/O receipt
from SSV-1 to Host at Site A
7.	 Inbound write I/O is destaged from
cache to internal storage
Challenge #1: Eliminate Single Points of Failure and Downtime
Data availability involves achieving data redundancy and maintaining accessibility even
in the event of a catastrophic failure within the storage architecture due to hardware
malfunctions, site failures, regional disasters, and user errors. DataCoreVirtual SAN provides
full data redundancy and eliminates interruptions using synchronous mirroring.
Synchronous Mirroring Order of Operations
3
High-speed Caching
The first feature responsible for storage acceleration within the
cluster is high-speed cache. Many other solutions today require the
use of flash as a caching device. DataCoreVirtual SAN leverages RAM
as high-speed cache, along with the hosts’ local processors and I/O
resources, to execute sophisticated multi-threaded read and write
caching algorithms across all storage devices under its management.
High-speed caching drastically improves overall application
performance by pre-fetching and storing the most frequently
accessed data in local memory, optimizing data transfers to
storage through write coalescing (sequential ordering), and
acting as a speed-matching write buffer to alleviate inbound
channel congestion. Up to 1Terabyte (TB) of RAM can be
reserved per node forming cluster-wide mega-caches.
Using RAM as a high-speed cache offers several unique
advantages since RAM is many orders of magnitude faster than
flash technologies. RAM is mirrored for inbound write data,
is not impacted by wearing or write-amplification effects like
flash, and resides closest to the CPU to minimize latency. It is the
fastest storage component in the storage architecture, helping
to deliver up to 10X performance boost to applications and
freeing up hosts to perform other tasks. It also extends the life of
traditional magnetic storage components by minimizing the stress
experienced from disk thrashing due to I/O intensive workloads.
While flash is inadequate to perform high-speed caching
functions, it can be used as persistent storage within a disk
pool and to accelerate frequently accessed data provided via
the auto-tiering feature. In this manner, its performance and
capacity attributes contribute to the system, as opposed to
other solutions which only use flash for a system-wide cache.
WHITEPAPER
Challenge #2: Improve Application Performance
Today’s demanding applications, especially within virtualized environments, require high
performance from storage to keep up with the rate of data acquisition and unpredictable
demands of enterprise workloads. DataCoreVirtual SAN provides substantial storage
acceleration through two primary features: high-speed caching & automated storage tiering.
4
Automated StorageTiering
The second feature responsible for storage acceleration within
the cluster is automated storage tiering. Automated storage
tiering is designed to move data to higher performing storage
tiers based on access frequency.When a high-capacity disk
is combined with high-speed storage devices such as flash,
it yields markedly better application performance at scale.
Unlike other auto-tiering solutions, which tend to be manual
or at best scheduled, DataCoreVirtual SAN auto-tiering is a
fully-automated real-time mechanism. In the background, it
analyzes data block requests and chooses the best placement
for your data based on access frequency. Both read and write
access patterns can influence the auto-tiering mechanism.
Also, auto-tiering policies apply on a per- virtual disk basis
granting you granular control over optimizations.
Automated storage tiering provides seamless integration of
high-speed storage devices, such as optional flash, with different
types of magnetic high-capacity devices (such as SAS and
SATA drives). A very compelling reason to use auto-tiering is
that only a small amount of flash disk is needed to significantly
improve application performance and user experience. Adding
just a small amount of flash, typically 8-10% of the total storage
pool capacity, yields significant performance enhancements.
DataCoreVirtual SAN supports many different tiers of
storage technology including cloud storage for infrequently
accessed, long-term archival data. As more advanced
technologies become available, tiers can be modified or
added as necessary to maintain performance service levels.
CPU-Storage Proximity
An important characteristic of DataCoreVirtual SAN is
that high-speed storage services reside as close to the
CPU and applications as possible.This type of architecture
reduces the number of components and the distance
involved in accessing the data.The result is greatly
improved application responsiveness and performance.
In general, RAM is the fastest and closest component to the
CPU. It is followed by PCIe flash, storage connected via an
internal controller, and finally external storage arrays. By
minimizing the number of components that must be traversed
to access the data, context switching is minimized, and
therefore latency. DataCoreVirtual SAN uses RAM as high-
speed cache, providing the closest proximity to storage services.
The locality of persistent storage is also an important design
consideration. As shown above, latency increases the further the
data resides from the CPU. Leveraging PCIe flash with internal
disks eliminates the inherent latency attributed with traversing
the external network or fabric.This combined with auto-tiering
provides a highly-optimized, high-performance storage solution.
WHITEPAPER
5
DataCoreVirtual SAN uses thin provisioning which allocates
storage only when data is written resulting in an increase of storage
utilization to as high as 95%.Thin provisioning is accomplished
by presenting a standard storage volume which appears to have
been fully provisioned and allocated, when in fact nothing has
been allocated. Only when the application writes data are the
storage blocks granularly allocated for that particular volume.
Conversely, automatic zero-block reclamation constantly
analyzes the thin provisioned storage blocks looking for those
which have been zeroed out due to data removal by the host.
But what about scalability? Expanding storage resources
separately from compute resources is another important
advantage of DataCoreVirtual SAN over other solutions.
To scale storage, you can choose to:
A.	 Expand capacity by adding more internal
storage to existing hosts (scale up)
B.	 Add hosts which contribute more storage (scale out)
C.	 Tapping into external storage on a central
SAN and/or cloud storage
D.	 Supplement on-premise storage with cloud storage
Note: Other hyper-converged alternatives confine you to
options A and B.They do not permit access to capacity
in your SAN (option C) or cloud storage (option D)
To scale out processing power, you may:
•	 Expand compute resources by adding
more hosts without data storage
•	 Expand compute and storage resources
concurrently by adding hosts
•	 Substitute faster hosts
In all cases, you have full control over how best to scale compute
and storage resources independently within the cluster.
WHITEPAPER
Challenge #3: Increase Efficiency and Scalability
DataCoreVirtual SAN RAM automatically improves storage efficiency and utilization through
thin provisioning.With storage systems that do not utilize thin provisioning, storage is
allocated on a per-volume per-application basis.This results in much of the available storage
being wasted since the unused space on each volume is inaccessible by other applications.
The industry average storage utilization without thin provisioning is around 35%.
6
Ultimate Storage Convergence:
As noted earlier, the DataCoreVirtual SAN can also take
advantage of existing central SANs and cloud storage
for additional capacity without having to add extra
components to the cluster. Unify all your storage under
DataCore control and manage it all from the same console.
This gives you the best of both worlds.The fastest storage
is right next to your applications directly on the host, yet
you can access more capacity or a different class of storage
from the central SAN or cloud storage as needs arise.The
auto-tiering software can do this for you without manual
intervention. No other product offer this capability!
You may also choose to mirror certain virtual disks
between your hosts and the central SAN / cloud storage,
rather than between hosts. Or, replicate between a
DataCoreVirtual SAN at remote locations to a central
DataCore SAN at the main data center or to cloud
storage.The point is that applications clusters don’t
need to become isolated as others would have you do.
They can be fully integrated into the broader storage
infrastructure to harness all the assets available to you.
WHITEPAPER
7
Deploying a Converged, Flexible Architecture
Consolidate your architecture to one-third of its traditional
size and introduce your choice of storage components
necessary to deliver the performance and capacity your
applications require. And as conditions change over time, deploy
additional storage components to meet the new resource
demands, all without interruption or risk to your data.
Achieving High Performance at Lower Cost
The use of low cost RAM as high-speed cache gives you a
significant performance-cost advantage. Additionally, over
time it also reduces the number of device failures since
the backend storage systems are not being overwhelmed
with constant high I/O activity.The use of RAM along with
less stress on storage components results in a significantly
lower total operating cost. And unlike other solutions,
the use of flash is optional with DataCoreVirtual SAN.
Integrate DataCoreVirtual SAN with
Central SAN / Cloud Storage
Whether you have an existing central SAN / Cloud
Storage architecture or you need to expand to one in
the future, DataCoreVirtual SAN seamlessly unifies
both architectures and provides a single pane of glass
management interface for the entire environment.
OnlyTwo Nodes NeededTo Start
DataCoreVirtual SAN only requires two nodes to get
started and can be incremented one node at a time. Other
solutions need three or more nodes to be fully functional.
Cross Hypervisor Support
DataCoreVirtual SAN supports all industry hypervisors
such as vSphere, Hyper-V, XenServer, and KVM. Different
hypervisors offer different advantages. It may be optimal
to deploy one type of hypervisor for critical enterprise-class
workloads such as line-of-business applications or databases
and another type for non-critical workloads such as virtual
desktops.Whether deployed at headquarters, at a branch
office, or both, DataCoreVirtual SAN will provide the same
enterprise class storage and management services.
WHITEPAPER
Challenge #4: Reduce Infrastructure Costs
DataCoreVirtual SAN significantly lowers yourTCO by:
8
Applicable Use Cases
DataCoreVirtual SAN opens up many new possibilities
within the infrastructure. Below are some of the most
common use cases:
Databases
Database applications such as SQL, Oracle, and SAP run
best in a low latency high-throughput storage environment.
DataCoreVirtual SAN delivers low latency storage access by
providing storage services as close to the application as possible.
Additionally, high-speed caching functions leverage RAM, which
is the fastest storage component in the system.The result is
significantly improved database application performance.
Virtual Desktop (VDI) Deployments
Run more virtual desktops (bothVMware Horizon
and Citrix XenDesktop) on each hypervisor host
and scale them out across more servers without the
complexity or expense of an external SAN.
Compact Server Clusters at Remote
Sites and Branch Offices
Put the internal storage capacity of your application
servers to work as a shared resource while protecting
your data against outages. Industries such as retail,
hospitality, manufacturing, and healthcare benefit greatly
from converged highly-available architectures.
Summary of Benefits
DataCoreVirtual SAN provides many benefits to your
infrastructure such as:
Highest Performance
Highest storage performance through the use of
high-speed RAM as cache.
Achieve end-to-end hardware independence
Just as server virtualization decouples ‘computing’
from the server hardware, storage virtualization
decouples the ‘data’ from the storage hardware,
resulting in complete infrastructure-wide flexibility.
Low upfront capital expenditure
Add resources as performance and capacity demands increase.
Conclusion
DataCoreVirtual SAN is the next evolution in software-defined storage which provides
high-performance and highly-available shared storage pools using the disks and flash
storage in your application servers; and if necessary, integrate external storage.
DataCoreVirtual SAN will significantly improve application performance and responsiveness, eliminate
single points of failure, increase storage efficiency and utilization, and lower infrastructure costs. It
provides complete storage flexibility, leveraging any make and model of storage device, all while
providing blazing performance and the most advanced forms of data protection. And as conditions
change over time, expand and adapt to fulfill new demands without interrupting access to your data.
WHITEPAPER
For additional information, please visit www.datacore.com or email info@datacore.com
© 2014 DataCore Software Corporation. All Rights Reserved. DataCore, the DataCore logo and SANsymphony are trademarks or registered trademarks
of DataCore Software Corporation. All other products, services and company names mentioned herein may be trademarks of their respective owners.
0714
WHITEPAPER
DataCore Software
DataCore is a leader in software-defined storage.The company’s storage virtualization
software empowers organizations to seamlessly manage and scale their data storage
architectures, delivering massive performance gains at a fraction of the cost of solutions
offered by legacy storage hardware vendors. Backed by 10,000 customer sites around the
world, DataCore’s adaptive and self-learning and healing technology takes the pain out of
manual processes and helps deliver on the promise of the new software-defined data center
through its hardware agnostic architecture. For more information go to www.datacore.com

More Related Content

PDF
Virtual SAN - A Deep Dive into Converged Storage (technical whitepaper)
DataCore APAC
 
PPTX
NetApp & Storage fundamentals
Shashidhar Basavaraju
 
PPT
CDW: SAN vs. NAS
Spiceworks
 
DOCX
Storage Area Networks Unit 2 Notes
Sudarshan Dhondaley
 
PDF
San nas-
Saurabh Chauhan
 
PDF
Technical Report NetApp Clustered Data ONTAP 8.2: An Introduction
NetApp
 
DOCX
Storage Area Networks Unit 1 Notes
Sudarshan Dhondaley
 
DOCX
Block Level Storage Vs File Level Storage
Pradeep Jagan
 
Virtual SAN - A Deep Dive into Converged Storage (technical whitepaper)
DataCore APAC
 
NetApp & Storage fundamentals
Shashidhar Basavaraju
 
CDW: SAN vs. NAS
Spiceworks
 
Storage Area Networks Unit 2 Notes
Sudarshan Dhondaley
 
San nas-
Saurabh Chauhan
 
Technical Report NetApp Clustered Data ONTAP 8.2: An Introduction
NetApp
 
Storage Area Networks Unit 1 Notes
Sudarshan Dhondaley
 
Block Level Storage Vs File Level Storage
Pradeep Jagan
 

What's hot (20)

DOCX
How to choose a server for your data center's needs
IT Tech
 
PPTX
DAS RAID NAS SAN
Ghassen Smida
 
PDF
FAQ on Dedupe NetApp
Ashwin Pawar
 
PPTX
Basics of storage Technology
Lopamudra Das
 
PPT
Storage Technology Overview
nomathjobs
 
PDF
Net App Unified Storage Architecture
nburgett
 
PDF
Hitachi overview-brochure-hus-hnas-family
Hitachi Vantara
 
PDF
Big Data SSD Architecture: Digging Deep to Discover Where SSD Performance Pay...
Samsung Business USA
 
DOCX
Storage Area Networks Unit 3 Notes
Sudarshan Dhondaley
 
PDF
Qnap nas ts serie x71u-catalogo
Fernando Barrientos
 
PDF
Why is Virtualization Creating Storage Sprawl? By Storage Switzerland
INFINIDAT
 
PDF
1 rh storage - architecture whitepaper
Accenture
 
PPT
Efs
Industech
 
PDF
Netezza Deep Dives
Rush Shah
 
PPTX
Massive parallel processing database systems mpp
Diana Patricia Rey Cabra
 
PDF
Netapp Storage
Prime Infoserv
 
PDF
Big Data Glossary of terms
Kognitio
 
PPT
Exadata
Maged Ali
 
PPTX
IBM Pure Data System for Analytics (Netezza)
Girish Srivastava
 
PDF
Exascale storage
Abhijeet Upponi
 
How to choose a server for your data center's needs
IT Tech
 
DAS RAID NAS SAN
Ghassen Smida
 
FAQ on Dedupe NetApp
Ashwin Pawar
 
Basics of storage Technology
Lopamudra Das
 
Storage Technology Overview
nomathjobs
 
Net App Unified Storage Architecture
nburgett
 
Hitachi overview-brochure-hus-hnas-family
Hitachi Vantara
 
Big Data SSD Architecture: Digging Deep to Discover Where SSD Performance Pay...
Samsung Business USA
 
Storage Area Networks Unit 3 Notes
Sudarshan Dhondaley
 
Qnap nas ts serie x71u-catalogo
Fernando Barrientos
 
Why is Virtualization Creating Storage Sprawl? By Storage Switzerland
INFINIDAT
 
1 rh storage - architecture whitepaper
Accenture
 
Netezza Deep Dives
Rush Shah
 
Massive parallel processing database systems mpp
Diana Patricia Rey Cabra
 
Netapp Storage
Prime Infoserv
 
Big Data Glossary of terms
Kognitio
 
Exadata
Maged Ali
 
IBM Pure Data System for Analytics (Netezza)
Girish Srivastava
 
Exascale storage
Abhijeet Upponi
 
Ad

Similar to Virtual SAN- Deep Dive Into Converged Storage (20)

PDF
[IC Manage] Workspace Acceleration & Network Storage Reduction
Perforce
 
PDF
SAN vs NAS vs DAS: Decoding Data Storage Solutions
MaryJWilliams2
 
PDF
What is a Network-Attached-Storage device and how does it work?
MaryJWilliams2
 
PDF
What is Network Attached Storage Used for?.pdf
Enterprisenas
 
PDF
Positioning IBM Flex System 16 Gb Fibre Channel Fabric for Storage-Intensive ...
IBM India Smarter Computing
 
PDF
TechDay - Toronto 2016 - Hyperconvergence and OpenNebula
OpenNebula Project
 
PDF
Insiders Guide- Managing Storage Performance
DataCore Software
 
DOC
Virtual Memory vs Cache Memory
Ashik Iqbal
 
PDF
SSDs Deliver More at the Point-of-Processing
Samsung Business USA
 
PDF
PowerAlluxio
Chi-fan Chu
 
PPTX
Introduction to Enterprise Data Storage, Direct Attached Storage, Storage Ar...
ssuserec8a711
 
PDF
Whitepaper
Tshombe Hamilton
 
PPT
Fulcrum Group Storage And Storage Virtualization Presentation
Steve Meek
 
PDF
Sansymphony v10-psp1-new-features-overview
Patrick Tang
 
PDF
Whitepaper_Cassandra_Datastax_Final
Michele Hunter
 
PDF
OUTBOARD SERVERS
befittingpiffle91
 
PDF
Cloud storage infrastructures
NIKHILKUMAR SHARDOOR
 
PDF
Supercharge Your Storage
DataCore Software
 
PPT
Veritas Failover3
grogers1124
 
[IC Manage] Workspace Acceleration & Network Storage Reduction
Perforce
 
SAN vs NAS vs DAS: Decoding Data Storage Solutions
MaryJWilliams2
 
What is a Network-Attached-Storage device and how does it work?
MaryJWilliams2
 
What is Network Attached Storage Used for?.pdf
Enterprisenas
 
Positioning IBM Flex System 16 Gb Fibre Channel Fabric for Storage-Intensive ...
IBM India Smarter Computing
 
TechDay - Toronto 2016 - Hyperconvergence and OpenNebula
OpenNebula Project
 
Insiders Guide- Managing Storage Performance
DataCore Software
 
Virtual Memory vs Cache Memory
Ashik Iqbal
 
SSDs Deliver More at the Point-of-Processing
Samsung Business USA
 
PowerAlluxio
Chi-fan Chu
 
Introduction to Enterprise Data Storage, Direct Attached Storage, Storage Ar...
ssuserec8a711
 
Whitepaper
Tshombe Hamilton
 
Fulcrum Group Storage And Storage Virtualization Presentation
Steve Meek
 
Sansymphony v10-psp1-new-features-overview
Patrick Tang
 
Whitepaper_Cassandra_Datastax_Final
Michele Hunter
 
OUTBOARD SERVERS
befittingpiffle91
 
Cloud storage infrastructures
NIKHILKUMAR SHARDOOR
 
Supercharge Your Storage
DataCore Software
 
Veritas Failover3
grogers1124
 
Ad

More from DataCore Software (20)

PDF
Software-Defined Storage Accelerates Storage Cost Reduction and Service-Level...
DataCore Software
 
PDF
NVMe and Flash – Make Your Storage Great Again!
DataCore Software
 
PDF
Zero Downtime, Zero Touch Stretch Clusters from Software-Defined Storage
DataCore Software
 
PDF
From Disaster to Recovery: Preparing Your IT for the Unexpected
DataCore Software
 
PDF
How to Integrate Hyperconverged Systems with Existing SANs
DataCore Software
 
PDF
How to Avoid Disasters via Software-Defined Storage Replication & Site Recovery
DataCore Software
 
PDF
Cloud Infrastructure for Your Data Center
DataCore Software
 
PDF
Building a Highly Available Data Infrastructure
DataCore Software
 
PDF
TUI Case Study
DataCore Software
 
PDF
Thorntons Case Study
DataCore Software
 
PDF
Top 3 Challenges Impacting Your Data and How to Solve Them
DataCore Software
 
PDF
Business Continuity for Mission Critical Applications
DataCore Software
 
PDF
Dynamic Hyper-Converged Future Proof Your Data Center
DataCore Software
 
PDF
Community Health Network Delivers Unprecedented Availability for Critical Hea...
DataCore Software
 
PDF
Case Study: Mission Community Hospital
DataCore Software
 
PDF
Emergency Communication of Southern Oregon
DataCore Software
 
PPTX
DataCore At VMworld 2016
DataCore Software
 
PPTX
Integrating Hyper-converged Systems with Existing SANs
DataCore Software
 
PPT
Fighting the Hidden Costs of Data Storage
DataCore Software
 
PPTX
Can $0.08 Change your View of Storage?
DataCore Software
 
Software-Defined Storage Accelerates Storage Cost Reduction and Service-Level...
DataCore Software
 
NVMe and Flash – Make Your Storage Great Again!
DataCore Software
 
Zero Downtime, Zero Touch Stretch Clusters from Software-Defined Storage
DataCore Software
 
From Disaster to Recovery: Preparing Your IT for the Unexpected
DataCore Software
 
How to Integrate Hyperconverged Systems with Existing SANs
DataCore Software
 
How to Avoid Disasters via Software-Defined Storage Replication & Site Recovery
DataCore Software
 
Cloud Infrastructure for Your Data Center
DataCore Software
 
Building a Highly Available Data Infrastructure
DataCore Software
 
TUI Case Study
DataCore Software
 
Thorntons Case Study
DataCore Software
 
Top 3 Challenges Impacting Your Data and How to Solve Them
DataCore Software
 
Business Continuity for Mission Critical Applications
DataCore Software
 
Dynamic Hyper-Converged Future Proof Your Data Center
DataCore Software
 
Community Health Network Delivers Unprecedented Availability for Critical Hea...
DataCore Software
 
Case Study: Mission Community Hospital
DataCore Software
 
Emergency Communication of Southern Oregon
DataCore Software
 
DataCore At VMworld 2016
DataCore Software
 
Integrating Hyper-converged Systems with Existing SANs
DataCore Software
 
Fighting the Hidden Costs of Data Storage
DataCore Software
 
Can $0.08 Change your View of Storage?
DataCore Software
 

Recently uploaded (20)

PDF
Generating Union types w/ Static Analysis
K. Matthew Dupree
 
PDF
Key Features to Look for in Arizona App Development Services
Net-Craft.com
 
PPTX
ASSIGNMENT_1[1][1][1][1][1] (1) variables.pptx
kr2589474
 
PDF
vAdobe Premiere Pro 2025 (v25.2.3.004) Crack Pre-Activated Latest
imang66g
 
PPTX
classification of computer and basic part of digital computer
ravisinghrajpurohit3
 
PPTX
The-Dawn-of-AI-Reshaping-Our-World.pptxx
parthbhanushali307
 
PDF
WatchTraderHub - Watch Dealer software with inventory management and multi-ch...
WatchDealer Pavel
 
DOCX
Can You Build Dashboards Using Open Source Visualization Tool.docx
Varsha Nayak
 
PPTX
Web Testing.pptx528278vshbuqffqhhqiwnwuq
studylike474
 
PPTX
Odoo Integration Services by Candidroot Solutions
CandidRoot Solutions Private Limited
 
PDF
Adobe Illustrator Crack Full Download (Latest Version 2025) Pre-Activated
imang66g
 
PDF
Balancing Resource Capacity and Workloads with OnePlan – Avoid Overloading Te...
OnePlan Solutions
 
PDF
Enhancing Healthcare RPM Platforms with Contextual AI Integration
Cadabra Studio
 
PPTX
Role Of Python In Programing Language.pptx
jaykoshti048
 
PDF
Exploring AI Agents in Process Industries
amoreira6
 
PDF
49784907924775488180_LRN2959_Data_Pump_23ai.pdf
Abilash868456
 
PDF
Using licensed Data Loss Prevention (DLP) as a strategic proactive data secur...
Q-Advise
 
PDF
On Software Engineers' Productivity - Beyond Misleading Metrics
Romén Rodríguez-Gil
 
PDF
lesson-2-rules-of-netiquette.pdf.bshhsjdj
jasmenrojas249
 
PDF
An Experience-Based Look at AI Lead Generation Pricing, Features & B2B Results
Thomas albart
 
Generating Union types w/ Static Analysis
K. Matthew Dupree
 
Key Features to Look for in Arizona App Development Services
Net-Craft.com
 
ASSIGNMENT_1[1][1][1][1][1] (1) variables.pptx
kr2589474
 
vAdobe Premiere Pro 2025 (v25.2.3.004) Crack Pre-Activated Latest
imang66g
 
classification of computer and basic part of digital computer
ravisinghrajpurohit3
 
The-Dawn-of-AI-Reshaping-Our-World.pptxx
parthbhanushali307
 
WatchTraderHub - Watch Dealer software with inventory management and multi-ch...
WatchDealer Pavel
 
Can You Build Dashboards Using Open Source Visualization Tool.docx
Varsha Nayak
 
Web Testing.pptx528278vshbuqffqhhqiwnwuq
studylike474
 
Odoo Integration Services by Candidroot Solutions
CandidRoot Solutions Private Limited
 
Adobe Illustrator Crack Full Download (Latest Version 2025) Pre-Activated
imang66g
 
Balancing Resource Capacity and Workloads with OnePlan – Avoid Overloading Te...
OnePlan Solutions
 
Enhancing Healthcare RPM Platforms with Contextual AI Integration
Cadabra Studio
 
Role Of Python In Programing Language.pptx
jaykoshti048
 
Exploring AI Agents in Process Industries
amoreira6
 
49784907924775488180_LRN2959_Data_Pump_23ai.pdf
Abilash868456
 
Using licensed Data Loss Prevention (DLP) as a strategic proactive data secur...
Q-Advise
 
On Software Engineers' Productivity - Beyond Misleading Metrics
Romén Rodríguez-Gil
 
lesson-2-rules-of-netiquette.pdf.bshhsjdj
jasmenrojas249
 
An Experience-Based Look at AI Lead Generation Pricing, Features & B2B Results
Thomas albart
 

Virtual SAN- Deep Dive Into Converged Storage

  • 1. DataCore™Virtual SAN introduces the next evolution in Software-defined Storage (SDS) by creating high-performance and highly-available shared storage pools using the disks and flash storage in your servers. It addresses the requirements for fast and reliable access to storage across a cluster of servers at remote sites as well as in high-performance applications. DataCoreVirtual SAN virtualizes the local storage on two or more physical x86- 64 servers. It can leverage any combination of magnetic disks (SAS, SATA) and optionally flash, to provide persistent storage services as close to the application as possible without having to go out over the wire (network or fabric).Virtual disks provisioned from DataCoreVirtual SAN can also be shared across the cluster to support the dynamic migration and failover of applications between hosts. DataCoreVirtual SAN addresses the challenges that exist today within many IT organizations such as single points of failure, poor application performance (particularly within virtualized environments), low storage efficiency and utilization, and high infrastructure costs. Let’s explore how DataCoreVirtual SAN addresses each of these challenges with a proven, 10th generation software-defined storage solution, purpose- built for maximizing storage resources across the enterprise. Virtual SAN - A Deep Dive into Converged Storage DataCoreTechnical Whitepaper Virtual SAN Benefits Only 2 servers needed for High-Availability RAM for I/OAcceleration Flash is Optional 100M IOPS, 64PB with 64 Host Cluster Works with all major Hypervisors Integrate Converged, SAN, and Cloud Storage WHITEPAPER
  • 2. 2 Synchronous Mirroring The availability of business-critical data is extremely important for your organization. DataCoreVirtual SAN provides protection against catastrophic events; outages that fault-tolerance measures alone cannot adequately protect against. Synchronous mirroring provides data-level redundancy to maximize data availability and keep applications running without interruption. DataCoreVirtual SAN accomplishes high-availability within the cluster by synchronously mirroring every data block between two fully active cluster nodes, where others may require three or more hosts to accomplish that same capability.The two nodes maintain redundant copies of the data ensuring that even if one cluster node becomes inoperable, the applications will not experience any interruption, and the data is always available. Synchronous mirroring also plays a role in providing redundant write cache.When running in a mirrored configuration, a high- speed write cache is enabled.Writes are only confirmed when both nodes acknowledge they have received the data. If a partner node becomes unavailable due to maintenance for example, then write data is still safe and destaged to persistent storage. Unlike other solutions, where performance degrades significantly when a loss of redundancy occurs, DataCoreVirtual SAN read cache and write-coalescing functions are still enabled, providing read acceleration and optimized destaging operations. Below is a diagram that outlines the order of operations that occurs during synchronous mirroring: WHITEPAPER 1. Inbound write I/O received from Host at Site A 2. Inbound write I/O committed to high-speed RAM cache on SSV-1 3. Inbound write I/O transferred to SSV-2 4. Inbound write I/O committed to high-speed RAM cache on SSV-2 5. Acknowledgement of write I/O receipt from SSV-2 to SSV-1 6. Acknowledgement of write I/O receipt from SSV-1 to Host at Site A 7. Inbound write I/O is destaged from cache to internal storage Challenge #1: Eliminate Single Points of Failure and Downtime Data availability involves achieving data redundancy and maintaining accessibility even in the event of a catastrophic failure within the storage architecture due to hardware malfunctions, site failures, regional disasters, and user errors. DataCoreVirtual SAN provides full data redundancy and eliminates interruptions using synchronous mirroring. Synchronous Mirroring Order of Operations
  • 3. 3 High-speed Caching The first feature responsible for storage acceleration within the cluster is high-speed cache. Many other solutions today require the use of flash as a caching device. DataCoreVirtual SAN leverages RAM as high-speed cache, along with the hosts’ local processors and I/O resources, to execute sophisticated multi-threaded read and write caching algorithms across all storage devices under its management. High-speed caching drastically improves overall application performance by pre-fetching and storing the most frequently accessed data in local memory, optimizing data transfers to storage through write coalescing (sequential ordering), and acting as a speed-matching write buffer to alleviate inbound channel congestion. Up to 1Terabyte (TB) of RAM can be reserved per node forming cluster-wide mega-caches. Using RAM as a high-speed cache offers several unique advantages since RAM is many orders of magnitude faster than flash technologies. RAM is mirrored for inbound write data, is not impacted by wearing or write-amplification effects like flash, and resides closest to the CPU to minimize latency. It is the fastest storage component in the storage architecture, helping to deliver up to 10X performance boost to applications and freeing up hosts to perform other tasks. It also extends the life of traditional magnetic storage components by minimizing the stress experienced from disk thrashing due to I/O intensive workloads. While flash is inadequate to perform high-speed caching functions, it can be used as persistent storage within a disk pool and to accelerate frequently accessed data provided via the auto-tiering feature. In this manner, its performance and capacity attributes contribute to the system, as opposed to other solutions which only use flash for a system-wide cache. WHITEPAPER Challenge #2: Improve Application Performance Today’s demanding applications, especially within virtualized environments, require high performance from storage to keep up with the rate of data acquisition and unpredictable demands of enterprise workloads. DataCoreVirtual SAN provides substantial storage acceleration through two primary features: high-speed caching & automated storage tiering.
  • 4. 4 Automated StorageTiering The second feature responsible for storage acceleration within the cluster is automated storage tiering. Automated storage tiering is designed to move data to higher performing storage tiers based on access frequency.When a high-capacity disk is combined with high-speed storage devices such as flash, it yields markedly better application performance at scale. Unlike other auto-tiering solutions, which tend to be manual or at best scheduled, DataCoreVirtual SAN auto-tiering is a fully-automated real-time mechanism. In the background, it analyzes data block requests and chooses the best placement for your data based on access frequency. Both read and write access patterns can influence the auto-tiering mechanism. Also, auto-tiering policies apply on a per- virtual disk basis granting you granular control over optimizations. Automated storage tiering provides seamless integration of high-speed storage devices, such as optional flash, with different types of magnetic high-capacity devices (such as SAS and SATA drives). A very compelling reason to use auto-tiering is that only a small amount of flash disk is needed to significantly improve application performance and user experience. Adding just a small amount of flash, typically 8-10% of the total storage pool capacity, yields significant performance enhancements. DataCoreVirtual SAN supports many different tiers of storage technology including cloud storage for infrequently accessed, long-term archival data. As more advanced technologies become available, tiers can be modified or added as necessary to maintain performance service levels. CPU-Storage Proximity An important characteristic of DataCoreVirtual SAN is that high-speed storage services reside as close to the CPU and applications as possible.This type of architecture reduces the number of components and the distance involved in accessing the data.The result is greatly improved application responsiveness and performance. In general, RAM is the fastest and closest component to the CPU. It is followed by PCIe flash, storage connected via an internal controller, and finally external storage arrays. By minimizing the number of components that must be traversed to access the data, context switching is minimized, and therefore latency. DataCoreVirtual SAN uses RAM as high- speed cache, providing the closest proximity to storage services. The locality of persistent storage is also an important design consideration. As shown above, latency increases the further the data resides from the CPU. Leveraging PCIe flash with internal disks eliminates the inherent latency attributed with traversing the external network or fabric.This combined with auto-tiering provides a highly-optimized, high-performance storage solution. WHITEPAPER
  • 5. 5 DataCoreVirtual SAN uses thin provisioning which allocates storage only when data is written resulting in an increase of storage utilization to as high as 95%.Thin provisioning is accomplished by presenting a standard storage volume which appears to have been fully provisioned and allocated, when in fact nothing has been allocated. Only when the application writes data are the storage blocks granularly allocated for that particular volume. Conversely, automatic zero-block reclamation constantly analyzes the thin provisioned storage blocks looking for those which have been zeroed out due to data removal by the host. But what about scalability? Expanding storage resources separately from compute resources is another important advantage of DataCoreVirtual SAN over other solutions. To scale storage, you can choose to: A. Expand capacity by adding more internal storage to existing hosts (scale up) B. Add hosts which contribute more storage (scale out) C. Tapping into external storage on a central SAN and/or cloud storage D. Supplement on-premise storage with cloud storage Note: Other hyper-converged alternatives confine you to options A and B.They do not permit access to capacity in your SAN (option C) or cloud storage (option D) To scale out processing power, you may: • Expand compute resources by adding more hosts without data storage • Expand compute and storage resources concurrently by adding hosts • Substitute faster hosts In all cases, you have full control over how best to scale compute and storage resources independently within the cluster. WHITEPAPER Challenge #3: Increase Efficiency and Scalability DataCoreVirtual SAN RAM automatically improves storage efficiency and utilization through thin provisioning.With storage systems that do not utilize thin provisioning, storage is allocated on a per-volume per-application basis.This results in much of the available storage being wasted since the unused space on each volume is inaccessible by other applications. The industry average storage utilization without thin provisioning is around 35%.
  • 6. 6 Ultimate Storage Convergence: As noted earlier, the DataCoreVirtual SAN can also take advantage of existing central SANs and cloud storage for additional capacity without having to add extra components to the cluster. Unify all your storage under DataCore control and manage it all from the same console. This gives you the best of both worlds.The fastest storage is right next to your applications directly on the host, yet you can access more capacity or a different class of storage from the central SAN or cloud storage as needs arise.The auto-tiering software can do this for you without manual intervention. No other product offer this capability! You may also choose to mirror certain virtual disks between your hosts and the central SAN / cloud storage, rather than between hosts. Or, replicate between a DataCoreVirtual SAN at remote locations to a central DataCore SAN at the main data center or to cloud storage.The point is that applications clusters don’t need to become isolated as others would have you do. They can be fully integrated into the broader storage infrastructure to harness all the assets available to you. WHITEPAPER
  • 7. 7 Deploying a Converged, Flexible Architecture Consolidate your architecture to one-third of its traditional size and introduce your choice of storage components necessary to deliver the performance and capacity your applications require. And as conditions change over time, deploy additional storage components to meet the new resource demands, all without interruption or risk to your data. Achieving High Performance at Lower Cost The use of low cost RAM as high-speed cache gives you a significant performance-cost advantage. Additionally, over time it also reduces the number of device failures since the backend storage systems are not being overwhelmed with constant high I/O activity.The use of RAM along with less stress on storage components results in a significantly lower total operating cost. And unlike other solutions, the use of flash is optional with DataCoreVirtual SAN. Integrate DataCoreVirtual SAN with Central SAN / Cloud Storage Whether you have an existing central SAN / Cloud Storage architecture or you need to expand to one in the future, DataCoreVirtual SAN seamlessly unifies both architectures and provides a single pane of glass management interface for the entire environment. OnlyTwo Nodes NeededTo Start DataCoreVirtual SAN only requires two nodes to get started and can be incremented one node at a time. Other solutions need three or more nodes to be fully functional. Cross Hypervisor Support DataCoreVirtual SAN supports all industry hypervisors such as vSphere, Hyper-V, XenServer, and KVM. Different hypervisors offer different advantages. It may be optimal to deploy one type of hypervisor for critical enterprise-class workloads such as line-of-business applications or databases and another type for non-critical workloads such as virtual desktops.Whether deployed at headquarters, at a branch office, or both, DataCoreVirtual SAN will provide the same enterprise class storage and management services. WHITEPAPER Challenge #4: Reduce Infrastructure Costs DataCoreVirtual SAN significantly lowers yourTCO by:
  • 8. 8 Applicable Use Cases DataCoreVirtual SAN opens up many new possibilities within the infrastructure. Below are some of the most common use cases: Databases Database applications such as SQL, Oracle, and SAP run best in a low latency high-throughput storage environment. DataCoreVirtual SAN delivers low latency storage access by providing storage services as close to the application as possible. Additionally, high-speed caching functions leverage RAM, which is the fastest storage component in the system.The result is significantly improved database application performance. Virtual Desktop (VDI) Deployments Run more virtual desktops (bothVMware Horizon and Citrix XenDesktop) on each hypervisor host and scale them out across more servers without the complexity or expense of an external SAN. Compact Server Clusters at Remote Sites and Branch Offices Put the internal storage capacity of your application servers to work as a shared resource while protecting your data against outages. Industries such as retail, hospitality, manufacturing, and healthcare benefit greatly from converged highly-available architectures. Summary of Benefits DataCoreVirtual SAN provides many benefits to your infrastructure such as: Highest Performance Highest storage performance through the use of high-speed RAM as cache. Achieve end-to-end hardware independence Just as server virtualization decouples ‘computing’ from the server hardware, storage virtualization decouples the ‘data’ from the storage hardware, resulting in complete infrastructure-wide flexibility. Low upfront capital expenditure Add resources as performance and capacity demands increase. Conclusion DataCoreVirtual SAN is the next evolution in software-defined storage which provides high-performance and highly-available shared storage pools using the disks and flash storage in your application servers; and if necessary, integrate external storage. DataCoreVirtual SAN will significantly improve application performance and responsiveness, eliminate single points of failure, increase storage efficiency and utilization, and lower infrastructure costs. It provides complete storage flexibility, leveraging any make and model of storage device, all while providing blazing performance and the most advanced forms of data protection. And as conditions change over time, expand and adapt to fulfill new demands without interrupting access to your data. WHITEPAPER
  • 9. For additional information, please visit www.datacore.com or email [email protected] © 2014 DataCore Software Corporation. All Rights Reserved. DataCore, the DataCore logo and SANsymphony are trademarks or registered trademarks of DataCore Software Corporation. All other products, services and company names mentioned herein may be trademarks of their respective owners. 0714 WHITEPAPER DataCore Software DataCore is a leader in software-defined storage.The company’s storage virtualization software empowers organizations to seamlessly manage and scale their data storage architectures, delivering massive performance gains at a fraction of the cost of solutions offered by legacy storage hardware vendors. Backed by 10,000 customer sites around the world, DataCore’s adaptive and self-learning and healing technology takes the pain out of manual processes and helps deliver on the promise of the new software-defined data center through its hardware agnostic architecture. For more information go to www.datacore.com