�ݺ�ߣ

Network Storage
Systems
1
Chapter 2

Router
Concentrator
Bridge
Dish
Network Switch/hub
Computer
System
HBA or
NIC
VPN
Connecting

 Networking or bus technology
 Cables + connectors
 System adapters + network device drivers
 Network devices such as hubs, switches, routers
 Virtual networking
 Flow control
 Network security
Connecting

Tape
Drives
Disk
Drives
RAID
Subsystem
Volume
Manager
Software
Mirroring
Software
(wiring, network transmission frame)
Storage Command and Transfer Protocol
Storage
Device
Drivers
Host
Software
Storage
Protocol
Storage
Devices
Storing

Device (target) command and control
 Drives, subsystems, device emulation
Block storage address space manipulation
(partition management)
 Mirroring
 RAID
 Striping
 Virtualization
 Concatentation
Storing

User/Application View
C:directoryfile
User/Application View
Database
Object
(Storage)
Logical Block Mapping
Filing

Connecting, storing and filing as a
complete storage system
Computer System
HBA or
NIC
Network Switch/hub
Disk Drive
Wiring Storing
Filing
Storing function in
an HBA driver
Cable Cable
Connecting

The classic storage model
The picture shown here represents the current state of conversations
about
storage networking: vendors, system designers, and customers
Things are made worse by there being a great many network storage
components, with relatively small differences between them.This
causes designs that are actually the same to be described in different
ways; and different designs to be described sometimes using
identical forms of words.This is clearly undesirable, and results in
many problems: it’s often not obvious what is being proposed or
described; tradeoffs between alternatives are harder to identify than
they could – or should –be; and it’s harder for everybody to make
high quality decisions.
9

These confusions are not accidental: the wide variety of the range of
system architectures that have been developed exhibit a great deal of
complexity because they are trying to accommodate a great deal of
information, and cover many different elements and functions. Some
of those elements are physical – boxes, wires,computers – and it is
often the case that architectures are presented by describing the
physical components in some detail, coupled with an explanation of
what functions they perform.That is, the traditional approach
focuses first on the physical partitioning that a particular vendor has
selected, rather than on the range of options that may be possible.
And because this is “box-centric” rather than “function-centric” it is
all too easy to misunderstand precisely what has been included.
11

The SNIA Shared Storage Model is an approach to removing these
difficulties. It does so by taking a slightly different approach: it
first identifies the functions that can be provided, and then
describes a range of different architectural choices for placing
those on physical resources.As a result, the SNIA Shared Storage
Model makes it easier to compare alternative architectures and
designs, it lets architects think about functions independently of
implementations, and it makes it simpler to anticipate new
implementations or combinations of architectures, designs, and
implementations.
12

SAS – ServerAttached Storage
Early mainframe storage designs took the premise that disk storage,
which was cheaper than main memory, could be treated as a
extended virtual memory to swap memory-pages.To achieve the
fast data access, the data paths (or channels)between storage and
processor were widened, the storage bus kept adjacent to the
processor bus for data/signal integrity while boosting the channel
speeds. Server attached storage architectures dominated the scene
for several years from mainframe processor channels to PC Server
bus slota and adapters.
14

One of the handicaps of the traditional server attached storage comes
from the tight coupling between storage and the operating system.A
general purpose SAS server performed a variety of tasks concurrently
from running applications,manipulating databases, file/print serving,
providing communications, checking data integrity to many
housekeeping functions.This meant that all data access requests
from a client must continuously compete with these asks
continuously.As the number of users accessing the common
centralized data storage increases, the file access takes a back seat to
other tasks leading to slow response time for queries. For years one of
the major jobs of MIS administrators was to keep the storage
performance fine tuned to achieve a certain minimum level of user
query response time.
16

Another limitation imposed by the SAS architecture was that of
limited distance imposed by the interface - the OEMI wide
parallel connections in mainframes and wide differential parallel SCSI
connections in servers were limiting the distance between
computers and servers to a few meters.This led to the creation of
raised-floor data centers but posed a severe constraint and
limitation on interconnectivity in multi-site operations. One of the
major benefits of fibre channel connectivity that is not fully
emphasized, is the removal of spaghetti of OEMI/SCSI wires
interconnecting storage to servers and the associated improvement
In reliability.This is over and above the advantage of allowing high-
speed connectivity and increased distance between centrally
managed data repositories and dispersed LAN servers.
17

NAS and SAN analysis
NAS is filing over a network
SAN is storing over a network
NAS and SAN are independent technologies
They can be implemented independently
They can co-exist in the same environment
They can both operate and provide services
to the same users/applications

Protocol analysis for NAS and SAN
Storing
Wiring
Filing
NAS
SAN
Network
Filing
Connecting
Storing

Storing
Wiring
Filing
NAS‘Head’
Server
System
NAS
Client
Wiring
SAN
Storage
NAS Server
+
SAN Initiator
“NAS Head”
SAN
Target
Connecting
Storing
Connecting
Filing
Integrated SAN/NAS environment

Storing
Wiring
Filing
NAS‘Head’
Server
System
NAS
Client
SAN
Storage
SAN
Target
Storing
Connecting
Filing
NAS Head
Common wiring with NAS and SAN

Network hardware path components
Cabling
Fiber optic
Copper
Switches, hubs, routers, bridges, gatways
Port buffers, processors
Backplane, bus, crossbar, mesh, memory

Routing
Fabric
Services
Virtual
Networking
Access and
Security
Flow
Control
Network software path components

24
 NAS – NetworkAttached Storage:
NetworkAttached Storage, compared to server attached storage on
the other hand is a dedicated file server optimized to do just one
function only and do it well - file serving.NAS is a system
independent, shareable storage that is connected directly to the
network and is accessible directly by any number of heterogeneous
clients or other servers. NAS file servers are essentially stripped
down servers specifically designed for file serving and offloading
file management services from the more expensive application
servers.

Factors motivating rise of NAS servers include:
26
Performance
Stored data supplied directly to clients without server intervention
Performance enhancements for a site can be achieved by
dedicating each NAS server for its specific needs
(e.g.Publishing/Prepress department can have its own file server
dedicated to video/imaging. graphics data using RAID-3 while the
e-commerce sales/order processing/shipping/customer service
groups could be running OLTP applications on its own dedicated
server, running RAID-5 or RAID-1attached to the same net.
Availability
Fault Resiliency -Majority of data has becoming mission critical
to run a business and so must be made secure and reliable Need
for 99.9% availability (8 Hours per year of downtime). Some
applications require even higher data availability such as 99.99%
(1 hour of downtime per year) and recovery from failure from
hardware, software and application switchover within 30 seconds.

Cost
HSM: migration to low cost tape for infrequently used data.
Scalability
Other benefits accruing to NAS architecture include modular
scalability by direct attachment of add-on file servers directly to
the net without bringing down any applications running already.
Interoperability
NAS is very capable of supporting heterogeneous clients (such as NT
and UNIX workstations) to share same data from network-attached
server. Majority of mainframe storage today acts as a physical
repository to store different types of file data such as UNIX/NFS
or MVS or NT/SMB in different physical locations of the storage
system.
27

Manageability
NAS lends itself to dedicated storage management resident on the
NAS servers itself to ensure efficient backups.
Challenges/Caveats
One of the major shortcomings of NAS storage architecture is that
the network on which NAS runs is also used for data access by
clients to retrieve data from the file server or communicate with
application servers.The data movement between the disk and tape
servers also goes over the same LAN.This creates a major network
bottleneck when the number of users increases.
28

SAN – Storage Area Network
 A SAN (Storage Area Network) is a dedicated high performance
network to move data between heterogeneous servers and storage
resources. Being a separate dedicated network it avoids any traffic
conflict between clients and servers.A fibre channel based SAN
combines the high performance of an I/O channel (IOPS and
bandwidth) and the connectivity (distance) of a network.
To interconnect distributed systems over distance IT system
administrators have been forced to use Fast Ethernet links,which
are terribly inefficient because of large packet overhead (associated
with small 1500 byte transmission packets) and high latency.
29

In smaller computer room environments, short, thick and unwieldy
spaghetti of SCSI wires or OEMI copper cables in mainframe
environments to connect storage to servers are commonplace.
Performance:SAN enables concurrent access of disk or tape
arrays by two or more servers at high speeds across fibrechannel,
providing much enhanced system performance.
Availability:SAN has disaster tolerance built in since data can be
mirrored using FC SAN up to 10 km away.
Cost:Since SAN is an independent network, initial costs to set up
the infrastructure would be higher but the potential exists for
rapid cost erosion as SAN installed base increases.
Manageability:Data centric ; Part of Server cluster;
Thin protocol for low latency;• DMA to server RAM - direct
communication to Data
30

summarizes the comparisons
between SAS, NAS and SANs
32

Quick Overview
33
DAS NAS SAN
Storage Type sectors
shared
files
blocks
Data
Transmission
IDE/SCSI
TCP/IP,
Ethernet
Fibre
Channel
Access Mode
clients or
servers
clients or
servers
servers
Capacity
(bytes)
109 109 - 1012 1012
Complexity Easy Moderate Difficult
Management
Cost (per GB)
High Moderate Low

SAN V/s NAS
Fibre Channel IP
Reliable Unreliable
Extremely low CPU Overhead Extremely High
Large Blocks of Data Large Nos. of small Blocks
LAN-free back-up Lan Back-up
Applications managing own
Data specs
Applications driven by universal
access to files

36
DAS NAS FC-SAN
FC
Switch
servers
storage
clients

SCSI Distribution Architecture
37
 SCSI is a client/server architecture.
 The client is called the initiator and issues request to the server. The
client is I/O subsystem under the typical OS control.
 The “server” is called the target, which is the SCSI controller inside
the storage device. It receives, process, and responds to the requests
from the initiator.
 SCSI commands support block I/O, transferring large amount
of data in blocks.
Client
(Initiator)
Storage Device
(Target)
request
response

SCSI Client/Server Architecture
38
Client
(Host)
Server
(Storage Device)

Redundant Array of Independent Disks (RAID)
 A group of hard disks is called a disk array
 RAID combines a disk array into a single virtual
device
 called RAID drive
 Provide fault tolerance for shared data and
applications
 Different implementations: Level 0-5
 Characteristics:
 Storage Capacity
 Speed: Fast Read and/or Fast Write
 Resilience in the face of device failure
39

Duplication
Parity
Difference
2n
n+1
-1
d(x) = f(x) – f(x-1) f(x-1)
RAID = parity redundancy

 RAID arrays use XOR for calculating parity
Operand 1 Operand 2 XOR Result
False False False
False True True
True False True
True True False
 XOR is the inverse of itself
 Apply XOR in the table above from right to left
 Apply XOR to any two columns to get the third
Parity redundancy

●Capacity scaling
● Combine multiple address spaces as a single virtual address
●Performance through parallelism
● Spread I/Os over multiple disk spindles
●Reliability/availability with redundancy
● Disk mirroring (striping to 2 disks)
● Parity RAID (striping to more than 2 disks)
Benefits of RAID

RAID Types
 RAID 0
 Stripe with no parity (see next slide for figure)
 RAID 1
 Mirror two or more disks
 RAID 0+1 (or 1+0)
 Stripe and Mirrors
 RAID 3
 Synchronous, Subdivided Block Access;
 RAID 5
 Like RAID 4, but parity striped across multiple drives
43

RAID 0 RAID 1
44
Disk Mirror
Disk Striping (no redundancy)

RAID 3 RAID 5
46
Disk striping with Dedicated Parity Drive

Network Attached Storage (NAS)
 NAS is adedicated storage device, and it operates in
a server mode.
 NAS is connected to the file server via LAN.
 Protocol: NFS (or CIFS) over an IP Network
 Network File System (NFS) – UNIX/Linux
 Common Internet File System (CIFS) – Windows Remote file
system (drives) mounted on the local system (drives)
 evolved from Microsoft NetBIOS, NetBIOS over TCP/IP (NBT), and Server
Message Block (SMB)
 SAMBA: SMB on Linux (Making Linux a Windows File Server)
 Advantage: no distance limitation
 Disadvantage: Speed and Latency
 Weakness: Security
47

 Specialized storage device or group of storage devices providing centralized fault-
tolerant data storage for a network
48
Clients
Servers Storage Devices

File Sharing Environment
 File system is structured way of storing and organizing data
files
 File Sharing
 Storing and accessing data files over network
 FS must be mounted in order to access files
 Traditional client/server model, implemented with file-sharing
protocols for remote file sharing
 Example: FTP, DNS
 DFS is another examples of file sharing implementation

File Sharing Technology Evolution
Stand Alone PC
Networked File Sharing
Networked PCs
Portable Media
for File Sharing

Components of NAS
Network Interface
NFS CIFS
NAS Device OS
Storage Interface
NAS Head
NFS
CIFS
UNIX
Windows
Storage Array
IP

NAS File Sharing Protocols
 Two common NAS file sharing protocols are:
 CIFS – Common Internet File System protocol
 Traditional Microsoft environment file sharing protocol, based upon the
Server Message Block protocol
 NFS – Network File System protocol
 Traditional UNIX environment file sharing protocol

Network File System (NFS)
 Client/server application
 Uses RPC mechanisms overTCP protocol
 Mount points grant access to remote hierarchical file
structures for local file system structures
 Access to the mount can be controlled by permissions

NAS File Sharing - CIFS
 Common Internet File System
 Developed by Microsoft in 1996
 An enhanced version of the Server Message Block (SMB)
protocol
 Stateful Protocol
 Can automatically restore connections and reopen files that were open
prior to interruption
 Operates at the Application/Presentation layer of the OSI
model
 Most commonly used with Microsoft operating systems, but is
platform-independent
 CIFS runs overTCP/IP and uses DNS (Domain Naming
Service) for name resolution

Benefits of NAS
 Support comprehensive access to information
 Improves efficiency
 Improved flexibility
 Centralizes storage
 Simplifies management
 Scalability
 High availability – through native clustering
 Provides security integration to environment (user
authentication and authorization)

Storage Area Network (SAN)
 A StorageArea Network (SAN) is a specialized, dedicated
high speed network joining servers and storage, including
disks, disk arrays, tapes, etc.
 Storage (data store) is separated from the processors (and
separated processing).
 High capacity, high availability, high scalability, ease of
configuration, ease of reconfiguration.
 Fiber Channel is the de facto SAN networking
architecture, although other network standards could be
used.
56

• The first major SAN networking technology
• Very low latency
• High reliability
• Fiber optic cables
• Copper cables
• Extended distance
• 1, 2 or 4 Gb transmission speeds
• Strongly typed
Fibre channel

A Fibre Channel fabric presents a consistent interface and set of services
across all switches in a network
Host and subsystems all 'see' the same resources
SAN
Target
Storage
Subsystem
SAN
Target
Storage
Subsystem
SAN
Target
Storage
Subsystem
Fibre channel

Ethernet / TCP / IP SAN technologies
 Leveraging the install base of Ethernet andTCP/IP
networks
 iSCSI – native SAN over IP
 FC/IP – FC SAN extensions over IP

Star
• Simplest
• single hop
 Dual star
• Simple network
+ redundancy
• Single hop
• Independent or integrated
fabric(s)
SAN topologies

N-wide star
• Scalable
• Single hop
• Independent or integrated
fabric(s)
 Core - edge
• Scalable
• 1 – 3 hops
• integrated fabric
SAN topologies

Ring
• Scalable
• 1 to N÷2 hops
 Ring + Star
• Scalable
• 1 to 3 hops
SAN topologies

SAN Benefits
 Storage consolidation
 Data sharing
 Non-disruptive scalability for growth
 Improved backup and recovery
 Tape pooling
 LAN-free and server-free data movement
 High performance
 High availability server clustering
 Data integrity
 Disaster tolerance
 Ease of data migration
 Cost-effectives (total cost of ownership)
63

NAS vs. SAN ?
 Traditionally:
 NAS is used for low-volume access to a large amount of storage by
many users
 SAN is the solution for terabytes (1012) of storage and multiple,
simultaneous access to streaming audio/video.
 The lines are becoming blurred between the two technologies
now, and while the SAN-versus-NAS debate continues, the fact is
that both technologies complement each another.
64

�ݺ�ߣ

Advanced DB chapter 2.pdf

More Related Content

Advanced DB chapter 2.pdf