Storage Concepts Student Guide
User Manual:
Open the PDF directly: View PDF 
.
Page Count: 262 [warning: Documents this large are best viewed by clicking the View PDF Link!]
Storage Concepts
Introduction
Student Introductions
•Name
•Position
•Experience
•Your expectations
Welcome and Introductions
This 2 day instructor-led course provides a comprehensive introduction
to storage technology concepts, terminology and technologies of
today’s storage industry. The course examines the need for storage
solutions to manage and optimize an IT infrastructure to meet business
requirements.
The course also examines the major components of a storage system,
common storage architectures and the various means of connecting
storage elements. It compares network attached storage (NAS) and
storage area network (SAN) implementations and data protection
issues. It provides detail on industry-defined tiered storage,
virtualization, and storage management strategies.
Course Description
Understanding of basic computer concepts
Experience working with PCs or servers (Windows or UNIX)
Prerequisites
By completing the course, you will gain an understanding of:
•Storage industry concepts and technologies
•Industry-defined tiered storage, virtualization, and storage management
strategies
Course Objectives
Course Topics
Modules Activities
1.
Introduction to Data Management
and Storage Systems
1a.Overview of Storage Concepts
2.
Storage Components and
technologies
Learning activities appear t
hroughout
the course.
3.
Business Continuity and
Replication
4.
Virtualization of Storage Systems
5.
Archiving and File and Content
Management
6.
Storage System Administration
7.
Business Challenges
8.
Storage Networking and Security
(Optional)
Storage Concepts
Introduction to Data Management and
Storage Systems
Module Objectives
Upon completion of this module, you should be able to:
•Explain the different types of data
•Explain what Cloud Computing is
• Explain a storage systems’ view of data
•Distinguish between the physical and logical levels of data processing
•Understand and explain the basic concepts of data consistency and data
integrity
Module Topics
Introduction to Data Management – Data Types
Structured and Unstructured Data
Data versus Information
Data Processing Levels
A Storage Systems View of Data
Data Consistency and Data Integrity Concepts and Principles
Introduction to Data
Management
Common Users: Data Examples
Photos
Movies
Documents
Email
Personal web pages
Data backed up to online
storage such as Microsoft Sky
Drive
Increasingly popular cloud
systems and on-line
applications
Movies
Accounting, invoices, and financial records
Databases that contain data about clients
Email communication
Digitalization of printed documents
Archiving
Business Sector Data
Databases
Audio and video records
Hospital records
Confidential and classified data
Archiving and digitalization
State Institutions Data
Medical
Records
Data Lifecycle applies to all data that comes into existence
•Data Retention Period applies to certain types of data and is governed by
law
Data Lifecycle
Offline
Structured Data
•Databases
Unstructured Data
•Medical Images – MRI scans
•Photographs
•Digital documents – check images
•Satellite images
•Biotechnology
•Digital video
•Email
Structured and Unstructured Data
Email
The Changing Forms of Data
New data types and business models compound the problem of exploding
storage. Consider the demands brought on by these new business models as
a result of the ubiquity of the Internet.
Total Digital Archive Capacity, by Content Type – Worldwide (TB)
30,000,000
2005 2006 2007 2008 2009 2010
25,000,000
20,000,000
15,000,000
10,000,000
5,000,000
0
Database
Unstructured Data
ESG Research Report: Digital Archiving: End-
User Survey & Market Forecast 2006-2010
Data
Information
Question: What’s The Difference Between Data and
Information?
Data Versus Information
Data
•A physical and written
representation of information
and knowledge
•Succession of written
characters, which can be
represented by numbers,
letters, or symbols
Information
•Meaningful interpretation of
data
•Does not have to be written in
characters
Data is stored to preserve and
pass information on
Data in binary code…
New Trends in Data
Management –
Cloud Computing
http://www.youtube.com/watch?v=ae_DKNwK_ms&feature=yout
u.be
Video – What Is Cloud Computing?
YouTube video
Three Key Cloud Characteristics
The 3 key characteristics of a cloud are:
Self-service
Pay-per-use
Dynamic scale up and down
“Cloud is a way of using technology, not a technology in
itself – it's a self-service, on-demand pay-per-use model.
Consolidation, virtualization and automation strategies
will be the catalysts behind cloud adoption.”
– The 451 Group
Levels of Data
Processing – Logical
And Physical
For example, a laptop running Microsoft Windows
software:
•Installed Hard Drives — CD / DVD / memory card drives
•Each drive assigned a letter = valid address within the Operating System
For example, hard drives have C: & D: addresses, DVD drive has
E: address
Hard Drive Partitioning (Physical Level)
A user may want to encrypt a partition that will contain critical data
Other reasons:
Exercise: Can You Think of Some Reasons for
Partitioning a Hard Drive?
You can also send your answers using the CHAT
Logical Level – Volumes / File Systems
Volume – logical interface
used by an operating system
to access data stored on a
particular media while using a
single instance of a file
system
File System – the way in
which files are named,
organized, and stored on the
hard drive
Summary – File System
Stores information about where data is
physically located
Stores metadata, containing additional
information
Maintains data integrity and allows users to
set access restrictions and permissions
Installed on a homogenous storage area
called a volume
Applications access a file system using an
application programming interface (API)
Sets up the way files are named and
organized
How the NTFS Works
Logical Units on a Storage System
A storage system is partitions of logical units striped across a large
number of hard drives
Logical Devices
Windows
UNIX
Logical Unit Number (LUN) Concept
A LUN is a logical device mapped to a storage port
Logical Devices
Windows
UNIX
LUNS are mapped
to servers
Some Definitions
Microcode: built-in software that works on the lowest layer of
instructions, directly controlling hardware equipment
•Most basic software – no graphical user interface (GUI)
•Types:
Microcode stored on individual hard drives
Microcode on a storage system also containing an integrated interface,
either a command line interface (CLI) or a GUI – firmware
Firmware: contains microcode and some kind of user interface
(menus, Icons)
Microcode and Firmware
Data Consistency: means you have valid, usable and readable data
•Point-in-time consistency: data is consistently as it was at any single
instant in time
For example, synchronous (continuous) data replication
• Transaction consistency: preventing “lost” transactions
•Application consistency
Data Consistency
Data Integrity: describes accuracy, reliability and correctness in
terms of security and authorized access to a file
•Policies: containing rules that govern access, preventing possible data
alterations
•Permissions and restrictions of access tools
Data Integrity
Module Summary
Upon completion of this module, you should have learned to:
•Explain the different types of data
•Explain what Cloud Computing is
• Explain a storage systems’ view of data
•Distinguish between the physical and logical levels of data processing
•Understand and explain the basic concepts of data consistency and data
integrity
Storage Concepts
Differentiate between common basic
storage architectures
Module Objectives
Upon completion of this module, you should be able to:
•Differentiate between common basic storage architectures
Storage Architecture Connectivity
File System
Application
Storage
Direct Attach
Storage
DAS
Parallel SCSI
or
Serial SCSI (FC)
File System
Application
Storage
Area
Network
Storage
Storage Area
Network
SAN
Application
Local
Area
Network
File System
Storage
Network Attach
Storage
NAS
Direct Attached Storage
Storage is directly attached to the server
No other device on the network can access
the stored data
Example of DAS – A PC with an attached
external disk drive, another; A Mainframe
direct connect to SAN storage array
Best used for accessing personal data or
high speed non-shared access
File System
Application
Storage
Direct Attach
Storage
DAS
Network Attached Storage
File oriented data access
Optimized for file serving
Easy Installation and Monitoring
No server intervention (or layer) required for
data access
Low Total Cost of Ownership
Use existing Network/cabling
Multiple protocol support (file sharing) using
NFS, SMB, CIFS, HTTP, etc.
NAS Heads
NAS Blades (New HDS-G-Series 400-800)
NAS Filers (HDS-HNAS 3000/4000 Series)
Application
Local
Area
Network
File System
Storage
Network Attach
Storage
NAS
Network Attached Storage
Targets midsize customers and remote, branch
offices of large organizations
Ideal for customers with a collaborative
environment that requires sharing of files
such as project management teams, law
offices, and design firms
–File serving
–Software development
–CAD/CAM
–Rich media
–Publishing and broadcast
–Archiving
–Near-line Data Storage to meet regulatory
requirements
Local
Area
Network
NAS
Server
Storage
Introduction to SAN
Storage Area Network
•Is a separate network that includes computer servers, disks, and other storage
devices
•Allows networking concepts to be applied to a server/storage model
•Has its own connections rather than using a fixed backbone network
•Has connections that utilize Fibre Channel equipment
•Allows very fast access among servers and storage resources
•Enables many servers to share many storage devices
•Designed for very high speed shared data storage up to 16 million nodes
Storage Area Network
Designed to attach computer storage devices
such as disk array controllers and tape libraries
to servers
Primary purpose is the transfer of data between
computer systems and storage elements
(switches/ directors [large switches])
Consists of a communication infrastructure and
a management layer so that data transfer is
secure
and robust
SAN and NAS can co-exist on the same network
infrastructure
Complicated (relatively) and expensive to
implement
Storage
Area
Network
SAN
Server
Storage
Storage Area Network
Best suited for complex data centers that
require high availability, scalability, reliability,
and performance
•Storage hosting providers
•International organizations that have multiple
data centers
•Businesses that implement Service Level
Agreements (SLAs)
•Disaster Recovery and, or Business Continuity
requirements
Storage
Area
Network
SAN
Server
Storage
Summary
Direct Attached Storage (DAS) – Storage is directly attached to
the application or file server
•Only one computer can access the storage
•A PC with an externally-attached hard disk drive
Network Attached Storage (NAS) – Multiple computers can
access and share the storage devices
•Accessed over an IP network
•Ideal for collaborative file sharing
Storage Area Networks (SAN) – Designed to attach computer
storage devices to servers
•Complex communication infrastructure organizes the connections,
storage elements, and computer systems so that data transfer is
secure and robust
•Best suited for:
Storage hosting providers
International organizations that have multiple data centers
Anyone with need for high speed shared data access
Storage Concepts
Storage Components and Technologies
Module Objectives
Upon completion of this module, you should be able to:
•Identify and describe the major components of a storage system
•Explain the different RAID levels and configurations
•Describe the midrange storage system architecture and its components
•Explain the factors directly influencing the performance of a storage
system
•Understand and explain the differences between a midrange and an
enterprise storage system
Module Topics
How a hard drive works
What means of hardware redundancy we have
How to describe a midrange storage system architecture and
components
Module Flow
Hard Drive
and how it
works
Hard drive
connectivity
Factors
influencing
performance
Other
storage
system
components
Overview of Disk
Array Components
A hard disk drive (HDD) is a nonvolatile storage device that stores
data on a magnetic disk.
Key components of a disk drive:
Disk Drive Components and Connectivity
Disk Drive Components and Connectivity
Spindle – A spindle holds one or more platters. It is connected to a
motor that spins the platters at constant revolutions per minute
(RPM)
Platter – A platter is the disk that stores the magnetic patterns. It is
made from a nonmagnetic material, usually glass, aluminum, or
ceramic, and has a thin coating of magnetic material on both sides.
A platter can spin at a speed of 7,200 to 18,000 RPM. The cost of an
HDD increases for a higher speed.
Disk Drive Components and Connectivity
Head – The read-write head of an HDD reads data from and writes
data to the platters. It detects (when reading) and modifies (when
writing) the magnetization of the material immediately underneath it.
Information is written to the platter as it rotates at high speed past the
selected head.
There is one head for each magnetic platter surface on the spindle;
these are mounted on a common actuator arm.
Actuator – An actuator arm moves the heads in an arc across the
spinning platters, allowing each head to access the entire data area,
similar to the action of the pick-up arm of a record player.
Disk Drive Components and Connectivity
The performance of an HDD is measured using the following
parameters:
•Capacity – The number of bytes an HDD can store. The current
maximum capacity of an HDD is 4TB.
•Data transfer rate – The amount of digital data that can be moved to or
from the disk within a given time. It is dependent on the performance of
the HDD assembly and the bandwidth of the data path.
The average data transfer rate ranges between 50-300 MB per
second.
•Seek time – The time the HDD takes to locate a particular piece of data.
The average seek time ranges from 3 to 9 milliseconds.
Transfer Rates – Performance
Disk Drive Components and Connectivity
HDD Bus
HDD Connectivity
Interfaces
Disk Drive Components and Connectivity
Bus Cables connect the storage central
processing unit (CPU) to the HDD
interface.
Interface is a device that enables the
connection of electrical circuits together.
Interfaces use the following standards:
•Parallel advanced technology attachment
(PATA)
•Serial advanced technology attachment
(SATA)
•Small computer systems interface
(SCSI)
•Serial attached SCSI (SAS)
Disk Drive Components and Connectivity
PATA
PATA is a standard used to connect HDDs to computers, based on
parallel signaling technology. PATA cables are bulky and can be a
maximum of 18 inches long, so they can be used only in internal
drives.
SATA
SATA evolved from PATA. It uses serial signaling technology. SATA is
a standard used to control and transfer data from a server or storage
appliance to a client application. Compared to PATA, SATA has the
following advantages:
•Greater bandwidth
•Faster data transfer rates – up to 600GB/sec
•Easy to set up and route in smaller computers
•Low power consumption
•Hot-swap support
SATA does not perform as well as SAS
Disk Drive Components and Connectivity
SCSI
A parallel interface standard used to transfer data between
devices on both internal and external computer buses.
SCSI advantages over PATA and SATA:
•Faster data speeds
•Multiple devices can connect to a single port
•Device independence; can be used with most SCSI compatible hardware
SCSI has the following disadvantages:
•SCSI interfaces do not always conform to industry standards.
•SCSI is more expensive than PATA and SATA.
Disk Drive Components and Connectivity
Disk Drive Components and Connectivity
Serial Attached SCSI (SAS)
Serial Attached SCSI has evolved from the previous SCSI standards
as it uses serial signaling technology. SAS is a standard used to
control and transfer data with SCSI commands from a server or
storage appliance to a client application.
SAS advantages over SCSI:
•Greater bandwidth
•Faster data transfer rates
•Easy set up and routing in smaller computers
•Low power
SAS cables are similar to SATA cables.
Redundant Array of
Independent Drives
Redundant Array of Inexpensive/Independent Drives (RAID) – A
method of storing data on multiple disks by combining various
physical disks into a single logical unit. A logical disk is a
combination of physical disks
RAID provides the following advantages:
Data consistency and integrity (security, protection from corruption)
Fault tolerance
Capacity
Reliability
Better Speed
Different types of RAID can be implemented, according to application
requirements.
RAID
RAID
The different types of RAID implementation are known as RAID
levels.
Common RAID levels:
•RAID-0
•RAID-1
•RAID-1+ 0
•RAID-5
•RAID-6
RAID-0 (Data Striping)
RAID-0
•RAID-0 implements striping; data is
spread evenly across two or more
disks.
RAID-0 Benefits:
•Easy to implement
•Increased performance in terms of
data access (more disks equals
more heads, which enables parallel
access to more data records)
RAID-0 Disadvantage:
• This RAID level has no redundancy
and no fault tolerance. If any disk
fails, data on the remaining disks
cannot be retrieved, which is the
major disadvantage. RAID-0 uses only data striping.
RAID-1 (Data Mirroring)
RAID-1
•Implements mirroring to create exact
copies of the data on two or more
disks.
RAID-1 Advantages:
•Reduces the overhead of managing
multiple disks and tracking the data.
•Read time is fast because the system
can read from either disk.
•If a disk fails, RAID-1 ensures there is
an exact copy of the data on the
second disk.
RAID-1 Disadvantages:
•The storage capacity is only half of
the actual capacity as data is written
twice.
•RAID-1 is expensive; doubles the
storage.
RAID-1+
RAID-1+
•RAID-1+0 is an example of multiple or nested RAID levels.
•A nested RAID level combines the features of multiple RAID levels. The
sequence in which they are implemented determines the naming of the
nested RAID level.
•For example, if RAID-0 is implemented before RAID-1, the RAID level is
called RAID-0+1. RAID-1+0 combines the features of RAID-0 and RAID-1
by mirroring a striped array.
RAID-1+ has the following advantages:
•Easy to implement
•Fast read/write speed
•Data protection
RAID-1+ has the disadvantage of high cost to implement.
RAID-1+
RAID-5
RAID-5
•RAID-5 consists of a minimum of three disks (two data and one parity)
•RAID-5 distributes parity information across all disks to minimize potential
bottlenecks if one disk fails, in which case parity data from the other disks
is used to recreate the missing information.
RAID-5 has the following advantages:
•The most common and secure RAID level
•Fast read speed
•Ensures data recovery if a disk failure occurs
RAID-5 has the following disadvantages:
•Extra overhead required to calculate and track parity data
•Slower writes because it has to calculate parity before writing data
RAID-5
RAID-6
RAID-6
•RAID-6 is similar to RAID-5 with an additional parity disk
•In RAID-5, if a second disk fails before the first failed disk has been
rebuilt, data can be irretrievably lost. The additional parity drive in RAID-6
provides a solution to this problem.
RAID-6 is designed for large environments and offers the following
benefits:
•Provides protection against double-disk failure
•Has a fast read speed
RAID-6 has the following disadvantages:
•Similar to RAID-5 plus the cost of the extra parity disk
•Slight performance overhead
RAID-6
RAID Type Configuration and Usage
Correction Copy – Occurs when a drive in a RAID group fails and a
compatible spare drive exists. Data is then reconstructed on the
spare drive.
Spare Disks — Sparing
Dynamic Sparing – occurs if the online verification process (built-in
diagnostic) determines that the number of errors has exceeded the
specified threshold of a disk in a RAID group. Data is then moved to
the spare disk, which is a much faster process than data
reconstruction
Spare Disks — Sparing
Correction Copy Parameters
Copy Back
No Copy Back
Exercise: RAID Configuration Options
Considering the following data storage needs, which RAID options would
provide the best performance?
1. Online transactional database (banking, stock market)
where performance and reliability is key
2. Search engine, or catalog system for a library
3. Overnight billing and inventory system
If vILT class, write your answers on blank lines
Building a Midrange
Storage System –
Components
A typical expansion unit or disk enclosure (based on SAS architecture),
consists of following components:
•Individual hard drives (SSD, SAS, SATA)
•Expander (buses and wires for connecting drives together)
•Power supplies
•Cooling systems (fans)
•Chassis
Expansion Unit / Disk Enclosure
Expansion Unit Connectivity
Each expansion unit is connected to both controllers
A midrange storage system contains main controller boards that are
equipped with components that can be put into three categories:
•Front end (connection to hosts or other storage systems)
•Cache (works as a buffer, has major influence on performance)
•Back end (connections to hard drive enclosures, RAID operations)
Back-End Architecture – Midrange Storage System
Back-End SAS Architecture Example
Cache
Cache
Cache is a temporary storage area for frequently used data. A
system can access the cached copy instead of the data in the
original location. This reduces the time taken to access data.
Cache Operations
The storage system’s microcode contains
algorithms that should anticipate what data is
advantageous to keep (for faster read access)
and what data should be erased. There are two
basic algorithms that affect the way cache is
freed up:
Least Recently Used (LRU) — Data that is
stored in cache and has not been accessed for
a given period of time (i.e., data that is not
accessed frequently enough) is erased.
Most Recently Used (MRU) — Data accessed
most recently is erased. This is based on the
assumption that recently used data may not be
requested for a while.
LRU
MRU
LRU
Queue
Cache Mirroring
Cache Data Protection
x GB
Controller 0
Read/Write Cache
x GB
Controller 1
Mirrored Write Cache
x GB
Controller 0
Mirrored Write Cache
x GB
Controller 1
Read/Write Cache
Controller 0
x GB Cache
Controller 1
x GB Cache
Interface Board – Ports connecting storage system to servers
Protocols
•Fibre Channel (FC) – Defines a multi-layered architecture for moving
data; allows data transmission over twisted pair and over fiber optic
cables
•Fibre Channel over Ethernet (FCoE) – Encapsulation of Fibre Channel
frames over Ethernet networks
•iSCSI – Interface connecting storage system to the LAN; allows
organizations to utilize their existing TCP/IP network infrastructure without
investing in expensive Fibre Channel switches
Front-End Architecture
Front-End Architecture – Other Components
In the frontend, we have QE8 FC port controllers that are part of the interface board.
These controllers are mainly responsible for conversion of FC transfer protocol into
PCIe bus used for internal interconnection of all components.. Notice the CPU and
local RAM memory (not cache).
In the event of a path failure, LUN ownership does not change. Data is
transferred via the backup path to CTL1 and then internally to CTL0,
bypassing CTL0 front end ports. This diminishes performance because
of internal communication.
Active-Passive Architecture
Symmetric Active-Active Controllers
Multiple paths to a single LUN
are possible. LUN ownership
automatically changes in the
event of path failure.
Unlike active-passive
architecture, active-active
architecture offers equal
access to the particular LUN
via both paths. This means the
performance is not influenced
by what path is currently used.
Controller Load Balancing – Active-active Architecture
No need to configure LUN ownership manually. Communication goes
either via CTL0 or CTL1. In the event of path failure, no bypassing is
necessary; therefore there is no communication overhead.
Fibre Channel Ports and their Configuration
Storage connected to an
External port for virtualization
Storage systems
connected via an
Initiator port for
replication
Host connected to a Target port
iSCSI Interface
Mixed SAN showing Fibre
Channel SAN connection for
production servers and an
iSCSI LAN for the
Test/Development servers
Test/Development
Production
Module Summary
Upon completion of this module, you should have learned to:
•Identify and describe the major components of a storage system
•Explain the different RAID levels and configurations
•Describe the midrange storage system architecture and its components
•Explain the factors directly influencing the performance of a storage
system
•Understand and explain the differences between a midrange and an
enterprise storage system
Storage Concepts
Business Continuity and Replication
Module Objectives
Upon completion of this module, you should be able to:
•Describe basic concepts of business continuity and replication
•Understand business impact analysis and risk assessment
•Describe basic concepts of disaster recovery
Module Topics
Business Continuity concepts
Business Impact Analysis and Risk Assessment
Back-up strategies and their implementation
How Business Continuity and Disaster Recovery connect to IT
Replication options
Concepts of clusters and geoclusters
Business Continuity
Concepts
Exercise: Data Center Disaster
Tornado destroys ABC
Corporation’s data center
•What needs to be prepared
ahead of time?
•Impacts on the business?
Write your answers on blank lines or send your answers via the CHAT
Set up a Business Continuity team
Exercise: Data Center Disaster
Tornado destroys ABC
Corporation’s data center
•What needs to be prepared
ahead of time?
•Impacts on the business?
Set up a Business Continuity team
A Disaster Recovery Plan
Another data center location
Suppliers ready to replace damaged equipment
Employees trained in responding to these situations
Staff in place at the alternate site to take over
Failback to main data center
Financial loss
Damaged reputation and company image
Identifies critical business processes and establishes rules and
procedures
•Implementation governed by regulations and standards
Business Continuity Management: Regulatory
Requirements
Sarbanes Oxley –
Corporate reporting
and financial results;
tells CEO and CFO
that they must be able
to defend accuracy of
their books
Basel II –
International banking
regulation that deals
with amortizing cost of
risk into financial
markets
British Standard For
Business Continuity
Management
(BS25999) – guidance
for determining
business processes
and their importance
North American
Business
Continuity
Standard –
(NFPA1600)
The implementation of business continuity concepts with respect to
the particular organization. The planning normally includes:
•Business Impact Analysis: Identifies which business processes, users
and applications are critical to the survival of the business
•Risk Assessment : Determines probability of threats to an organization
• Policies: Aligns BC policies with the company’s business strategy
•Business Recovery Plan: Defines procedures to be taken when a
particular situation occurs
•Disaster Recovery Plan: Describes how to get the critical applications
working again after an incident takes place
•Testing and Training Schedules: Provides a timeline for business
continuity plan testing
Business Continuity Planning
Recovery Point Objective (RPO) – Worst case time between last
backup and interruption time
•Represents how much data must be recovered
•How much can you afford to lose?
Recovery Time Objective (RTO) – How long is the customer willing
to live with downed systems?
•Represents outage duration
Business Impact Analysis – RPO and RTO
Disaster Recovery – Part of business continuity that focuses only on IT
Infrastructure
Disaster Recovery Plan contains:
Basic information – purpose, area of application, requirements, log of
DR plan modifications, members of DR team, their roles and
responsibilities
Notification/activation phase – notification procedure, call tree,
damage assessment, activation criteria, plan activation
Recovery procedures – succession of recovery procedures according
to their importance, logging, escalation
Standard operation resumption – checking whether all systems work
properly, termination of DR plan
Amendments – call book, vendor SLA, RTO of processes
Business Continuity versus Disaster Recovery
Data Backup and
Data Replication
Concepts
Is data backup different from data replication?
Exercise: Data Backup and Data Replication
Write your answers on blank lines or send your answers via the CHAT
Backup =
Replication =
Example: Data Backup Configuration
Back-up over
LAN is the
simplest solution.
The back-up
server pulls data
from production
servers and then
it sends data to a
tape library or
NAS device.
Data Backup Models
Full Backup: Data is stored in
exact copies
Incremental Backup: Only the
data that was changed or added
since last time backup is
recorded
Differential Backup: Only the
data that differs from the initial
full backup is recorded
Reverse Delta Backup: At
every scheduled backup, the
initial full backup image is
synchronized so that it mirrors
the current state of data on
servers
To back-up the data from production servers you need:
A back-up device – in enterprise environment it will most probably be a
tape library, but it can also be a storage system with a LUN dedicated for
back-up.
A back-up server – in most cases you need a back-up server that is
communicating directly with the back-up device and that controls back-up
from all the servers
Back-up software – software that runs on a back-up server and that
allows to make configuration according to your needs
Back-up agents – these agents are small applications installed on all
your production servers. They are part of back-up software and they
allow communication between a back-up server and production servers.
Backup Requirements
Backup Optimization
Techniques used to achieve better backup utilization:
Compression – the output archive file is smaller than total of the
original files
Deduplication – the technique that eliminates duplicities in data
Multiplexing – the ability of software and equipment to back-up data
from several sources simultaneously
Staging – back up to a disk first and then transfer the data from this
disk to tape – known as Disk-to-disk-to-tape (D2D2T)
A volume with source data is called a Primary Volume (P-VOL), and
a volume to which the data is copied is a Secondary Volume (S-VOL).
In-system – all operations with logical units (LUs) within the same
storage system
Remote – all operations with LUs across different storage systems
Data Replication Overview
P-VOL S-VOL
Data Replication Overview
Data replication (or protection) provides operational and disaster recovery
•Replicates data within or between storage systems without disruption
•Creates multiple protected copies from each source volume
•Can run independent of host OS, database, file system
•Mirrors image of data
•Offers quick restart and recovery in disaster situations
Once created, copies can be used for:
•Data warehousing or data mining applications
•Backup and recovery
•Application development
Data Replication
Within and between array heterogeneous replication
In the event of a major disaster, would you use your backup solution
or data replication solution to recover your business critical online
applications?
•Break up into teams, discuss, and present your reasons for choosing one
of the solutions over the other.
Exercise: Data Backup Or Data Replication?
If a vILT class, present your answers over the phone
Data copy operations:
•Initial Copy
All data is copied from
P-VOL to S-VOL
Copies everything
including empty blocks
•Update Copy
Only differentials are
copied
Data Replication – Copy Operations
Requirements for All Replication Products
Any volumes involved in replication operations (source and
destination) should be:
•Same size (in blocks)
•Must be mapped to a port
Source can be online and in use.
Destination must not be in use/mounted.
Intermix of RAID levels and drive type is supported.
Licensing
•License is capacity independent.
P-VOLs S-VOLs
Primary Server Backup Server
Pairs
Create Pair
Updates
S-VOLs are inaccessible
after the paircreate
command is issued.
Paircreate
Data Replication Operations – Establish Pairs
Status change: SIMPLEX to COPY to PAIR
P-VOL S-VOL
Primary Server Backup Server
Status change: PAIR to PSUS
Pairsplit
Backup S-VOL
backup data
to tape
Updates are
marked in
differential
bitmaps
S-VOL now accessible.
Updates are marked in
S-VOL differential bitmaps
Data Replication Operations – Split Pairs
P-VOL S-VOL
Online
Primary Server Backup Server
Resynchronize
Pair
Updates
PAIRRESYNC
Normal Resync
Reverse Resync
Caution: Any changes applied
to S-VOL while the pair is split
are discarded during NORMAL
resync.
Status: PSUS to PAIR
Data Replication Operations – Resynchronizing Pairs
S-VOL P-VOL
Primary Server Backup Server
DUPLEX
Quick Restore Pair
Swaps LDEV Mapping
Updates
Quick Restore
Status: Split to PAIR with copy direction reversed
Data Replication Operations – Quick Restore
In-System
Replication
In-System Replication
In-system hardware-based copy facility that
provides:
•Full volume point-in-time copies
•Host Independent
•Nondisruptive replication
•Clone images are RAID protected copies
LU # 1
LU # 2
Production Data
(P-VOL)
Clone Image
(S-VOL)
Storage System
Snapshots
•Create a point-in-time
(PiT) copy or “snapshot”
of the data
•Uses less space than full
copies or clones
Frequent, cost effective,
point-in-time copies
•Multiple copies of a
primary volume
•Immediate read/write
access to virtual copy
•Fast restore from any
virtual copy
Copy on write snapshot. Notice that both P-VOL and V-VOL are
accessible for I/O operations. Snapshots can be created instantly
In-System Replication: Copy-On-Write Snapshots
In-System Replication: Copy-On-Write Snapshots
Copy-on Write virtual volume (V-VOL) maintains a view of the primary volume
(P-VOL) at a particular point in time
V-VOL is a composite of original data in the P-VOL and change data in the pool
V-VOL presents as a full volume copy to any secondary host
Since V-VOL does not copy all data, it can be created or deleted almost instantly
Remote Replication
Remote Replication Scheme
P-VOL S-VOL
Fibre
Channel Fibre
Channel
Extender Extender
DWDM / ATM / IP
Any Distance
Remote replication scheme; DWDM, ATM or IP
connections to remote site are possible.
Primary Host
Synchronous Remote Replication
The remote I/O is not posted “complete” to the
application until it is written to a remote
system
The remote copy is always a “mirror” image
Provides fast recovery with no data loss
Limited distance – response-time impact
2
3
1
4
P-VOL S-VOL
Provides a remote “mirror” of any customer data
•The remote copy is always identical to the local copy
•Allows very fast restart/recovery with no data loss
•No dependence on host operating system, database,
or file system
•Distance limit is variable, but typically less than 100
kilometers
• Impacts application response time
•Distance depends on application read/write activity,
network bandwidth, response-time tolerance and
other factors
Synchronous Replication
Asynchronous Remote Replication
1 2
P-VOL S-VOL
•The local I/O is disconnected from the
remote I/O
•Very little impact to response time over
any distance
•Data integrity and update sequence
maintained over any distance
•Fast restart/recovery
Fibre
Channel Fibre
Channel
3
Extender Extender
Dynamic Replication
Appliance
Dynamic Replication Appliance
A possible implementation of a Replication Appliance. Data is collected
from servers over LAN and then it is send to a storage system. Each
server is running an agent that splits the data.
Remote Replication and Geoclusters
Geocluster
interconnection scheme.
Both sites (local and
remote) are equipped
with the same nodes.
Data from the Disk Array
A are synchronously
replicated to Disk Array B
over iFCP, FCIP or Dark
Fiber technology, both
SANs are interconnected.
Servers in both locations
are also interconnected,
usually using TCP/IP
protocol.
.
Three data center multi-target replication. The maximum possible data
protection is ensured by using two remote sites for data replication.
Three Data Center Multi-target Replication
Diversity in Data Protection Requirements
Solution Area of Cost, Performance and Distance
Exercise: Replication Scenario
Scenario:
•Financial services business with two data centers 300 miles apart
Your task:
•Describe a disaster recovery strategy for this business using data replication.
If a vILT class, use your drawing tools to show your configuration on the slide
Module Summary
Upon completion of this module, you should have learned to:
•Describe basic concepts of business continuity and replication
•Understand business impact analysis and risk assessment
•Describe basic concepts of disaster recovery
Storage Concepts
Virtualization of Storage Systems
Module Objectives
Upon completion of this module, you should be able to:
•Understand and explain virtualization concepts and its benefits
•Explain the difference between fat and thin provisioning
•Describe SAN virtualization concepts
Module Topics
Virtualization concepts, features, and benefits
Different types of virtualization
“Fat” and “Thin” provisioning concepts and features
Virtualization
Concepts
What Is Virtualization?
Definition (source: www.wikipedia.org)
•Virtualization is the abstraction of computer resources.
•Hides the physical characteristics of computing resources from the way in
which other systems, applications, or end users interact with those
resources
•Single physical resources appear to function as multiple logical
resources, or multiple physical resources appear as a single logical
resource.
Virtualization has moved “out of the box” and into the infrastructure,
or cloud, and virtualization solutions are available at these layers:
•Virtualization of applications
•Virtualization of computers
•Virtualization of networks
•Virtualization of storage
The traditional architecture model requires one physical server
per operating system and application. A virtualized server is
able to run several virtual machines that all share the physical
hardware.
Server Virtualization
Virtualization
Elements
Elements of Virtualization
Virtualization Areas
Server virtualization is increasingly used; it provides better utilization
of server resources
One-to-many virtualization – Makes one physical server look like
many servers; allows multiple operating systems on one physical
server
Server Virtualization
Layers of Virtualization
Users can access the application
from s a virtual desktop.
Applications can run on several
virtual machines
The server can be virtualized
The host can access a virtualized
volume
VMware® Based Server Virtualization
The hypervisor virtualization layer is thin and optimized for direct access to
hardware resources. Virtual machines in child partitions access the virtualization
layer through the VMBus interface. Device drivers for virtual machines are loaded
from the parent partition with the original instance of Windows 2008 Server.
Virtual machine configuration and management are also done in the parent
partition operating system.
Hyper-V™ Based Server Virtualization
Blade servers are
installed in a blade
server chassis. The
chassis is then placed in
a standard rack. These
blade servers offer
logical partitioning,
which is a highly
sophisticated form of
server virtualization.
Virtualization with Blade Servers
Storage System Virtualization
Every storage system offers RAID functionality but additionally
may offer other virtualization capabilities such as:
Cache Partitions
Virtual Ports
Storage virtualization (of other storage systems)
Thin provisioning and automatic tiering
Virtualization Benefits
Migration – VMs and LUNs
can be easily transferred from
one physical device to another.
Backup – encapsulation
simplifies backup of the VM
Hardware platform
independence – physical
servers can be of different
configurations
Enhanced utilization – allows
effective use of resources
Lower power consumption
Lower RPO and RTO
Physical resources can be
added without disruption
Virtualized Server Cluster
Migration
Migration
Virtualization – Thin
Provisioning and
Automated Tiering
Purchased,
Allocated
BUT UNUSED
Actual DATA
Comparison of Fat and Thin Provisioning
To avoid future service interruptions, today it is common to over-
allocate storage by approximately 50% - 75%.
Thin Provisioning
Parity groups are
added to a thin
provisioning pool.
Virtual volumes
are mapped to
servers. Virtual
volumes do not
contain any actual
data. Data is stored
in the storage pool.
Virtual volumes
contain pointers that
point to the location
of data in the pool.
Thin Provisioning Benefits
•Increased physical disk striping for better performance
•Reduces the need for performance expertise
•Simplifies storage capacity planning and administration
•Increases storage utilization
•Eliminates downtime for application storage capacity
expansion
•Improves application uptime and SLAs
Zero Page Reclaim
The storage system scans the storage pool for used data
blocks that contain only zeros. These blocks are then erased
and freed automatically.
Thin Provisioning
An example of how thin provisioning can help you save the
cost of buying all the capacity in advance.
Exercise: Virtualization
1. Definition of Virtualization
•Virtualization is the ____________ of computer resources.
•Hides the __________________ of computing resources from the way in
which other systems, applications, or end users interact with those resources.
•Single physical resources appear to function as ____________________ .
•Or multiple physical resources appear as _____________________.
2. Identify some of the common objects that can be virtualized:
•___________________________
•___________________________
If a vILT class, write your answers on blank lines
Virtualization in
Storage System
Controller
Virtualization in Storage System Controller
Controller based virtualization of external storage. Hitachi
Virtual Storage Platform (VSP) is an example of an
enterprise-level storage system that supports virtualization
of external storage.
Before
virtualization the
data center
consists of
various
heterogeneous
storage systems
in a SAN.
Virtualization in Storage System Controller
After virtualization the VSP storage system provides access to
virtualized volumes from the other storage systems
Virtualization –
Logical Partitioning
FC/IP
SAN
Partition 1
Storage Virtualization – Logical Partitioning
In this figure we see one
storage system (VSP) with
three external storage
systems that create a
virtualized storage pool. The
VSP is then virtualized to
provide two logical partitions
— private virtual storage
machines. Hosts are then
able to access and use only
the resources (cache, ports
and disks) assigned to the
respective partition.
Exercise: Benefits
From your point of view, what do you think are the benefits of partitioning
storage?
If a vILT class, send your answers via the CHAT
or over the phone
Dynamic (Automated) Tiering
All the benefits of
Dynamic Provisioning
•Further simplified
management
•Further reduced
OPEX
•Better Performance
Dynamic Tiering
Dynamic
Provisioning
Dynamic Tiering in Hitachi Virtual Storage Platform
Storage systems are heading towards fully virtualized
solutions.
Cloud or Virtualized
Storage
The Future of Virtualization
Network
Storage
Direct Attached
Storage
Market Adoption Cycles
Cloud Storage
“Cloud is a way of using technology, not a technology in
itself – it is a self-service, on-demand pay-per-use
model. Consolidation, virtualization and automation
strategies will be the catalysts behind cloud adoption.”
– The 451 Group
Key characteristics of the cloud are:
The ability to scale and provision dynamically in a cost
efficient way
The ability to make the most of new and existing
infrastructure without having to manage the complexity of
the underlying technology
Cloud architecture can be:
•Private: Hosted within an organization’s firewall
•Public: Hosted on the internet
•Hybrid: A combination of private and public
Module Summary
Upon completion of this module, you should have learned to :
•Understand and explain virtualization concepts and its benefits
•Explain the difference between fat and thin provisioning
•Describe SAN virtualization concepts
Storage Concepts
Archiving and File and Content
Management
Module Objectives
Upon completion of this module, you should be able to:
•Explain the basic concepts and features of archiving
•Describe what is meant by Fixed Content
•Understand the differences between archiving and backup
Module Topics
Fixed Content and its characteristics
Components of a digital archive
Content management
Introduction to
Archiving
Fixed Content
What is Fixed Content?
•Data objects that have a long-term value, do not change over
time, and are easily accessible and secure
Email
Legal Records
Biotechnology
Digital Video
Medical Records
Satellite Images
Motivation for Data Archiving
There are several reasons for implementing data archiving
policies and technical solutions. Some of the most
prominent reasons are:
•Effective utilization of high performance tiers
•Cheaper storage for fixed content
•Data retention regulation
•Simplified content management
•Indexing and searching capabilities of a digital archive
Seeking an archiving solution. The storage systems we have discussed
up until now are not very suitable for fixed content storage and archiving.
Archive Solutions?
Many organizations have increasing regulations, especially in the
pharmaceutical industry, the food processing industry, healthcare, financial
services and auditing. The Sarbanes-Oxley Act very strictly regulates the
length of retention of financial records and accounting in companies.
Legal Requirements of Data Retention Periods
An example of a decentralized and fragmented archiving solution.
Disparate storage systems do not provide a common search engine,
and they are not very scalable. A digital archive can solve this problem.
The Need for a Better Archiving Solution
A Digital Archive
A digital archive works on the object level. Each object
contains fixed content data, metadata and description of
policies.
Block Level Storage Compared to Object Level Storage
A traditional block level storage system compared to an
object level storage system. The object level storage system
consists of powerful proprietary servers and management
software. These servers are connected to a RAID array.
Internal Object Representation
A data object and its components in detail. This example
illustrates how objects are handled by a object storage’s
digital archive.
Digital Archive Features
Active functions of a digital archive are:
•Content verification – Ensures authenticity and integrity of
each data object
•Protection service – Ensures stability of the digital archive
•Compression service – Achieves better utilization of
storage space assigned to the digital archive
•Deduplication service – Detects and removes duplicities
•Replication service – Ensures redundancy of archived
data
•Search capabilities – Allows users to search documents
Digital Archive Accessibility
In this example, a digital archive can be accessed using
multiple independent standard protocols. WebDAV is an
extension to HTTP protocol that allows remote management
of files stored on web server
Digital Archive Compliance Features
The most important compliance features of a digital archive
are:
Write once read many (WORM)
Retention period definition
Data shredding
Data encryption
Exercise: Fixed Content
Which 2 statements are true about the definition of fixed content?
(Choose 2)
Fixed content is …
a. Content that cannot be archived and restored
b. Content that can only be changed by the system administrator
c. Static data that is in a final state
d. Content that will not / cannot change
Answers= and
If vILT class, write your answers on blank lines,
Exercise: Backup versus Archive
Explain the differences between what is meant by Backup and Archive.
If vILT class, Send your answers to the instructor
via the WebEx CHAT tool or phone.
Exercise: Object Representation
Fixed-content data (Data)
System metadata
Custom metadata
1. Describe what each is.
2. What does an object contain?
Send your answers using CHAT
If vILT class, write your answers on blank lines
Upon completion of this module, you should have learned to:
•Explain the basic concepts and features of archiving
•Describe what is meant by Fixed Content
•Understand the differences between archiving and backup
Module Summary
Storage Concepts
Storage System Administration
Module Objectives
Upon completion of this module, you should be able to:
•Describe everyday storage administrator tasks
•Explain how to configure and monitor storage systems
•Describe tools used by the storage administrator in managing storage
Module Topics
Storage system administrator tasks and common functions
Storage system management software
Storage system implementer tasks
Storage
Administrator
In charge of maintaining a storage system infrastructure
•Tasks based on Service-level Agreements (SLAs)
Who is a Storage System Administrator?
Storage Administrator Tasks
Capacity Management
Amount of data to be stored; size
and performance of LUNs; hard drive
performance; I/O performance and
R/W operations
Availability Management
Replication, backup, and archive
strategies; protection against
component failures
Continuity Management
Part of business continuity planning
and disaster recovery procedures
Financial Management
Budget preparation; cost calculation
and invoicing and TCO
Storage Administrator Tasks – Other Common
Operations
Configuration of RAID
groups and volumes
Implementation of
changes in volume
configuration
Data replication
optimization
Configuration of cache
Cache partitioning
Backup of storage system
configuration
Integration of a New Storage System
When purchasing a new storage
system, the storage administrator
must think through the whole
implementation process, including
the following items:
Storage system model
Switch model
Cabling
Rack usage and floor space
Power requirements
Air conditioning
LAN infrastructure
Tasks of a Storage System Implementer
Installation and initial configuration of the storage system
Basic training to familiarize the customer with the new
device.
Conduct all hardware replacement and upgrade
procedures
Monitor the storage system remotely
Help with performance tuning
Microcode updates
Module Summary
Upon completion of this module, you should have learned to:
•Describe everyday storage administrator tasks
•Explain how to configure and monitor storage systems
•Describe tools used by the storage administrator in managing storage
Storage Concepts
Business Challenges
Module Objectives
Upon completion of this module, you should be able to:
•Identify business challenges driving the need for storage
•Understand why storage systems are important for business
• Explain the energy and green issues faced by today’s businesses
Module Topics
Business challenges companies face
Advanced data classification and tiered storage
Data center operations environmental concerns
Business
Challenges
Business Challenges
Business challenges can be classified as follows:
Accelerating storage growth — need more and more capacity
Increasing requirements on high availability — cannot afford any
disruptions because our data has become too important for our business
Fast and effective response to business growth — need to be able to
react quickly to new conditions
Heterogeneous infrastructure — result of fast infrastructure growth,
which was not properly planned, causing TCO to increase rapidly
Compliance and security challenges — need to process, protect and
retain data according to legal requirements and regulations
Power and cooling challenges — pay too much for electricity and
cooling and may be running out of space in the server room
Data center challenges — specific needs for those whose business is
focused primarily on cloud type provision of storage and computing
capacity
Data Growth Forecast
An overview of the storage requirements in the past years
provide you with the necessary information to forecast
data growth.
Data Growth Forecast – Tier View
Data growth forecast in relation to performance tiers
Structured, Unstructured and Replicated Data Growth
Advanced Data Classification and Tiering
Tiered storage infrastructure
Power and cooling exceeds server
Spending
$0
$10
$20
$30
$40
$50
$60
$70
$80
1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009
Installed base
(M units)
2
4
6
8
10
12
14
16
18
Spending
(US$B)
New server spending
Power and cooling
What is the greatest facility problem
with your primary data center?
Gartner, Best Practices in Data Center Facilities
Excessive
Heat Insufficient
Raised Floor
Insufficient
Power Poor
Location Excess
Facility Cost
None of
the above
29%
21%
29%
6% 3%
13%
N = 112
Power Requirements and Cooling
Greatest issues organizations and companies face
Power Challenges Facing Data Centers
Some of the challenges facing data centers with respect to
electricity, cooling and environmental requirements include:
Running out of power, cooling and space
Growing energy costs
Increasing regulatory compliance issues
Data center expansion without consideration for future power and cooling
requirements
Data storage configured without adequate consideration to heat
distribution (equipment racks should be installed with cold rows and hot
rows
Difficulty relieving data center hot spots without disrupting applications
Other Challenges Facing Data Centers
In addition to power consumption metrics in kW, there are
other metrics that should be considered:
Total five-year power and space costs
Heat loading (kW/sq ft)
Space requirements (sq ft)
Floor loading (lbs/sq ft)
Controller-based virtualization and thin provisioning can
also yield substantial environmental advantages because
they reduce the need for storage capacity.
HDD and Fan Power Savings
Features that work on the HDD and fan level and lead to
significant power savings:
Spin down drives in selected RAID groups
SATA drives will park heads when idle for more than
2 hours
Adjust fan speeds to maintain correct temperatures
Keep data in cache as long as possible
Green Data Center
Hot and Cold Rows
Arrange racks in alternating rows with cold air intakes
facing one way and hot air exhausts facing the other
Module Summary
Upon completion of this module, you should have learned to:
•Identify business challenges driving the need for storage
•Understand why storage systems are important for business
• Explain the energy and green issues faced by today’s businesses
Storage Concepts
Storage Networking and Security
Module Objectives
Upon completion of this module, you should be able to:
•Describe basic networking concepts
•Explain how common network devices operate
•Explain how devices communicate in a network
•Explain storage area network security
Module Topics
•Basic networking concepts
•Operations of common network devices
•Possibilities we have in storage system
networking
•How devices communicate with each other
through the network
•How to secure Storage Area Networks
Introduction to
Networks –
Components
Twisted Pair Cable structure, RJ45 connectors
Twisted Pair Cable
Fiber Optic Cable
Fiber Optic Cables – Fiber optics is a technology that uses glass or
plastic fibers to transmit data as light impulses. A fiber optic cable
consists of a bundle of fibers and each fiber can transmit millions of
messages modulated onto light waves.
Fiber Optic Connectors
Fiber Optic Cable LC connector and SFP transceiver
Network nodes are all the devices connected in the network. We
distinguish between endpoint communication nodes and data
redistribution nodes.
Storage Network Components – Nodes
Storage Network Components – Ports
Ports
On a storage network, a port enables a node to communicate with
another node over a Fibre Channel connection.
•A node can contain multiple ports.
On a storage network, a port enables
the following connections:
•Server to switch
•Switch to switch
•Switch to storage
Storage Network Components – HBAs
Host Bus Adapter (HBA) – In a storage system, an HBA is a Fibre
Channel interface card installed in a server. It connects a computer
and storage devices on a network.
Each HBA has a unique WWN. The two types of WWNs on an HBA
are these:
•Node WWN: Shared by all ports on an HBA
•Port WWN: Unique to each port on the HBA
Storage Network Components – WWNs
WWN address, World Wide Name, is a unique label, which identifies a
particular device in Fiber Channel network
WWN example - 5 0 0 6 0 E 8 0 1 0 4 5 3 0 3 0 1 6
The Open Systems Interconnection (OSI) Model
The Open Systems
Interconnection (OSI)
model is a conceptual
model that
characterizes and
standardizes the
internal functions of a
communications system
by partitioning it into
abstraction layers.
Hub – a simple device that allows interconnection and
communication among nodes.
Storage Network Components – HUB
Router – provides an interface between two different networks
Storage Network Components – Routers
Directors
A director is a large and complex switch. It is:
•Highly available, reliable, scalable, and manageable
•Fault tolerant with the ability to recover from a non-fatal error
•Designed with redundant hardware components
•Capable of supporting Fibre Channel and fiber
connectivity (FICON)
•Potentially expensive and complex
•Designed for enterprises with large data centers
Large networks often use Fibre Channel switches
and directors in the same implementation.
Storage Network Components – Directors
Exercise: Storage Network Components
Match the storage network component with the appropriate description:
a. Node
b. Port
c. WWN
d. HBA
e. Cable
1. Connects and transmits signals between nodes
2. Transmits or receives data over a network
3. Fibre Channel interface card
4. Enables a node to communicate with another
node
5. Unique number used to identify elements on a FC
storage network
If vILT class, write your answers on blank lines
Storage Networking
Topologies
SAN Topologies
Point-to-Point (FC-P2P) – A point-to-point (P-P) topology is considered
the simplest topology, in which two devices are directly connected using
Fibre Channel.
•It has fixed bandwidth; data is transmitted serially over a single cable.
•It can be used with DAS.
SAN Topologies
Arbitrated Loop (AL) – An FC topology where all devices are part of
a loop and only one device can communicate with another device at
a time
In AL, devices use an access request mechanism called arbitrate
(ARB), which circles the loop.
•A device can use ARB depending on its priority and access rights.
•The device with the highest priority gets first access.
SAN Topologies
Switched Fabric (FC-SW) – A Fibre Channel topology that connects
multiple devices by using Fibre Channel switches
In a switched fabric topology, bandwidth is not shared between
devices, enabling devices to transmit and receive data at full speed
at all times.
Direct Attached Storage
Direct attached storage infrastructure. Server is directly connected to
a storage system. Storage system can be accessed only through the
server. Server can be accessed from Local Area Network (LAN).
Storage Area
Network (SAN)
Storage Area Network
Storage Area Network (SAN) is a high-speed network of shared
storage devices. Servers attached to a SAN can access any SAN
attached storage devices.
The only components in a SAN are storage devices and switches.
SAN is designed to connect computer storage devices, such as disk
array controllers and tape libraries, to multiple servers, or hosts.
SAN Components
Storage Area Networks are using Fiber Channel
infrastructure, which includes Host Bus Adaptors installed
in servers, Fiber Channel cables and switches, Fiber
Channel ports installed in the storage system front end and
proprietary network protocols.
Host Bus Adaptor (HBA) and an example of WWN number.
SAN over iSCSI Interface
Internet Small Computer Systems Interface (iSCSI) – SAN can be
implemented by using a network protocol standard called iSCSI
which uses the SCSI protocol to transmit data over TCP/IP networks.
iSCSI allows organizations to use their existing TCP/IP network
infrastructure without investing in expensive Fibre Channel switches.
iSCSI HBA
Network Attached
Storage
Network Attached Storage
Network Attached Storage (NAS) is represented by the server that
functions as the NAS Head and common storage system. There are
solutions that integrate both these functionalities in one package (NAS
Appliances). NAS devices work relatively independently; they do not
require servers with applications. All clients, application and other servers
can access files stored in a NAS device.
File Access Protocols
supported include:
•CIFS
•NFS,
•HTTP
•HTTPS
•FTP
Network Attached Storage
In NAS, the storage device is directly connected to the LAN and
there is no server between the data and other network devices.
Data is presented to the server at file level.
LAN
Network Attached Storage
Advantages of NAS
•Offers storage to different open-systems operating systems over
LAN
•Data is presented to servers at file level, reducing server overhead
•Dedicated file server, optimized for sharing files between many
users
•Minimizes overhead by centrally managing storage
•Facilitates easy and inexpensive implementation
Disadvantages of NAS
•Relies on the client-server model for communication and data
transport which creates network overhead
•Lower performance than a SAN
NAS Implementations
Methods of NAS Implementation – An organization can implement
NAS architecture by following methods:
•NAS appliance or filer
•NAS blade
•NAS gateway
NAS Appliance
NAS Appliance – Combines a front-end file server and back-end
storage system in a single unit. This approach is called a closed-box
approach.
NAS appliance has the following advantages:
•Combines a file server with the storage array
•Provides efficient performance
•Has high reliability
•Enables easy installation, management, and use
•Provides the least expensive NAS implementation
NAS appliance has the following disadvantages:
•Is not scalable
•No pool storage, which makes it hard to achieve high utilization.
NAS Blade – Allows multi-protocol data storage in a large disk array
NAS blade has following advantages in addition to a NAS appliance:
•Is scalable
•Provides backup of storage data
•Supports multiple NAS blades
NAS Blade
NAS Gateway
NAS Gateway – All devices communicate directly with the file
system.
A NAS gateway overcomes the limitations of a NAS appliance.
It has the following advantages:
•Separates file server from storage device
•Is less expensive than a NAS appliance
•Supports multiple NAS gateways
•Has better utilization rates
•Combines NAS with SAN capacity to meet growing storage requirements
•Provides NAS functionalities to SAN storage
NAS gateway controller uses FC protocol to connect to external
storage.
Converged solution – SAN and NAS together
•NAS head with storage over the SAN
•NAS scales to the limits of the SAN
Limited by NAS file system’s capacity
Co-exists with application servers
Centrally managed
Converged Solution – SAN and NAS Together
LAN
Hosts
Users
SAN
Raid Storage
NAS Gateway
Storage Networking Architectures…Side by Side
IP Network
Application
SAN
Application
File System
File System
DAS NAS SAN
Direct
Connected
Exercise: Storage Networking Concepts
1. A DAS device is not shared, so no other network device can access the
data without first accessing the server. True or False?
2. Select the best description for the Fibre Channel topology known as
FC-AL.
a) Two devices, data transmitted serially over a single cable
b) Multiple devices connected in a loop, highest priority device gets first
access
c) Multiple devices connected using Fibre Channel switch, devices transmit
and receive data at full speed at all times
Network Protocols
Protocol
Protocol – A set of rules that govern communication between
computers on a network. It regulates the following characteristics of
a network:
•Access method
•Physical topologies allowed in the network
•Types of cable that can be used in the network
•Speed of data transfer
Protocols
The different types of protocols that can be used in a network are:
•Ethernet
•Fibre Channel protocol (FCP)
•Fiber connectivity (FICON)
•Internet protocol (IP)
•Internet small computer system interface (iSCSI)
•Fibre Channel over IP (FCIP)
•Internet Fibre Channel protocol (iFCP)
•Fibre Channel over Ethernet (FCoE)
Protocols
Ethernet
Uses an access method called carrier sense multiple access/collision
detection (CSM/CD). Before transmitting, a node checks whether any other
node is using the network. If clear, the node begins to transmit. Ethernet
allows data transmission over twisted pair or fiber optic cables and is mainly
used in LANs. There are various versions of Ethernet with various speed
specifications.
FCP
Defines a multi-layered architecture for moving data. FCP packages SCSI
commands into Fibre Channel frames ready for transmission. FCP also
allows data transmission over twisted pair and over fiber optic cables. It is
mainly used in large data centers for applications requiring high availability,
such as transaction processing and databases.
FICON
Connects a mainframe to its peripheral devices and disk array. Ficon is
based on FCP and has evolved from the older ESCON protocol.
Protocols
IP/TCP
IP is used to transfer data across a network. Each device on the network
has a unique IP address that identifies it. IP works in conjunction with the
TCP, iSCSI and FCIP protocols. When you transfer messages over a
network by using IP, IP breaks the message into smaller units called
packets (third layer in OSI model). Each packet is treated as an individual
unit. IP delivers the packets to the destination. TCP Is the protocol that
combines the packets into the correct order to reform the message that
was sent from the source.
iSCSI
Establishes and manages connection between IP-based storage devices,
and it hosts and enables deployment of IP-based storage area networks. It
facilitates data transfers over intranets, manages storage over long
distances and is cost-effective, robust and reliable. iSCSI is best-suited
for web server, email and departmental business applications in small to
medium sized businesses.
Protocols
FCIP
Fibre Channel over IP is a TCP/IP based tunnelling protocol that
connects geographically distributed Fibre Channel SANs. FCIP
encapsulates Fibre Channel frames into frames that comply with TCP/IP
standards. It can be useful when connecting two SAN networks over the
Internet tunnel, in a similar fashion to virtual private networks (VPNs)
allowing connection to a distant LAN over the Internet.
iFCP
iFCP is again TCP/IP based. It is basically an adaptation of FCIP using
routing instead of tunneling. It interconnects Fibre Channel storage
devices or SANs by using an IP infrastructure. iFCP moves Fibre
Channel data over IP networks by using iSCSI protocols.
Both FCIP and iFCP provide means to extend Fibre Channel networks
over distance. Both these protocols are highly reliable and scalable. They
are best suited for connecting two data centers for centralized data
management or disaster recovery.
Protocols
FCoE
Fibre Channel over Ethernet is an encapsulation of Fibre Channel frames
over Ethernet networks. This allows Fibre Channel to use 10Gb Ethernet
networks while preserving the Fibre Channel protocol. FCoE provides
these advantages:
Network (IP) and storage (SAN) data traffic can be consolidated using a
single network switch.
It reduces the number of network interface cards required to connect
disparate storage and IP networks.
Reduces the number of cables and switches.
Reduces power and cooling costs.
Thus, you can build your SAN using Ethernet cables (mostly twisted pair).
You can use one switch for your IP-based network traffic (LAN) and for
creating SAN infrastructure. Even though the switch and cabling are the
same, LAN will run on TCP/IP while SAN runs on FCP.
Exercise: Storage Networking
Match the following list of components with the appropriate definition:
a. Client-server
b. Protocol
c. LAN
d. WAN
1. Network connecting devices in a small
geographic area
2. Relationship between two computers – one
sends requests; one responds with data
3. Set of rules governing communication
among computers on a network
4. Network connecting devices across larger
geographical areas
If vILT class, write your answers on blank lines
Storage Area
Network Security
LUN Mapping
A LUN is a logical device mapped to a storage port.
Logical Devices
Windows
UNIX
LUNS are mapped
To servers
Zones – Defined to establish rules governing communication of
network devices
•WWN Zoning (Soft Zoning)
•Port Based Zoning (Hard Zoning)
•Mixed Zoning
Zoning
An example of WWN based zoning
Module Summary
Upon completion of this module, you should have learned to:
•Describe basic networking concepts
•Explain how common network devices operate
•Explain how devices communicate in a network
•Explain storage area network security
