Storage Basics

1
 Contents

 Understand Storage Basics

 In a Nutshell
 Introducing the Concepts
 Direct Attached Storage (DAS)
 Storage Area Networks (SAN)
 Network Attached Storage (NAS)
 NAS Gateways
 ISCSI

2
 Storage

 DAS
 SAN
 NAS
 NAS Gateways
 Fibre Channel
 iSCSI

3
The Data Flow Chart…

4
 Storage types
 Single Disk Drive
 JBOD
 Volume
 Storage Array – RAID’s
 SCSI device
 DAS
 NAS
 SAN
 iSCSI

5
Storage
 Storage Capacity
 RPM
 Data Transfer speed
 Ave Seek Time

6
RAID

7
RAID

8
RAID

9
RAID

10
 Internal & External Storage
Internal Storage  RAID controllers & disk drives are internal
Server
to the server
Storage  SCSI, ATA, or SATA protocol between
Disk Drive controller and disks
RAID Controller
Server
RAID Controller
 RAID controller is internal
 SCSI or SATA protocol between Storage
controller and disks
Disk Drives
 Disk drives are external

SCSI Bus w/ external storage

11
DAS w/external controller & Storage
Storage System
Server
Disk Enclosure

CPU
Disk Drives

Memory Disk Drives

Disk Drives
Bus

HBA

RAID
Controller

12
 DAS over Fibre Channel
 HBA is internal Server
 Fibre Channel protocol between HBA
HBAs and external RAID controller

Storage
 Disk drives and RAID controller are RAID Controller
external Disk Drives

External SAN Array

13
 I/O Transfer
 RAID Controller
 Contains the “smarts”
 Determines how the data will be written
(striping, mirroring, RAID 10, RAID 5, etc.)
 Host Bus Adapter (HBA)
 Simply transfers the data to the RAID controller.
 Doesn’t do any RAID or striping calculations.
 “Dumb” for speed.
 Required for external storage.

14
In a Nutshell….DAS

15
 DAS - SAN
 DAS : Direct Attached Storage
– Storage (usually disk or tape) is directly attached by a cable to the
computer processor.
– The hard disk drive inside a PC or a tape drive attached to a single
server are simple types of DAS.
– I/O requests (also called protocols or commands) access devices
directly.
 SAN : Storage Area Network
– SAN is dedicated network behind the servers , based on Fibre
Channel architecture

16
In a Nutshell….SAN

17
In a Nutshell….SAN
 SAN : Storage Area Network
– Storage resides on a dedicated
network.
– Like DAS, I/O requests access
devices directly.
– SANs use Fibre Channel media,
providing an any-to-any connection
for processors and storage on that
network.
– Ethernet media using an I/O protocol
called iSCSI is emerging in 2001.

18
 How SAN works
 The data from a host server is converted into optical light pulses by a “host bus
adapter” in the server, the pulses are transmitted over fiber-optic cables, through a
switched network, to an intelligent storage array, which uses advanced raid
protected disk drives to store the data safely.
 The act of using a network to create a shared pool of storage devices is what
makes a SAN different from the way data was normally stored on computers. The
network is used to move data between various storage devices, allows for data
sharing between different servers, and provides a fast medium for backing up and
restoring data.
 Devices in a SAN are usually bunched closely together in a single room, but the
network allows the devices to be connected over long distances. The ability to
spread everything out over long distances makes a SAN very useful to large
companies with many offices. 19
SAN implementation

20
 Who should buy SAN
 Database servers: Oracle, Sybase, SQL, DB2, Informix & other database servers
 File servers: Using SAN-based storage for file servers lets expand file server resources
quickly, makes them run better, and enables to manage file-based NAS storage through
the SAN.
 Backup servers: SAN-based backup is dramatically faster than LAN-based backup.
 Voice/video servers: Voice and video servers tend to push large amounts of data very
quickly.
 Mail servers: Using SAN-based storage for mail servers enables quick restoration of
data in case of corruption or viruses.
 High-performance application servers: Applications such as document management,
customer relationship management, billing, data warehouses, and other high-
performance and critical applications all benefit by what a SAN can provide.

21
 SAN Segments
 PARTS
– Host Layer- HBA, Drivers,
Pathing sofware, OS
– Fabric Layer- Hubs and
switches, fabric os, cabling
– Storage Layer- Tapes and disks,
advanced storage software
 PLAYERS :
– EMC, IBM,HITACHI,
NetApp,Sun, HP, Veritas etc

22
 SAN Components
H
B
A Fibre
C Cables
A
R
D

Tape Library

Storage Arrays

23
 SAN Topologies

 Point to Point
 Arbitrated Loop
 Switched Fabric

24
 Point to Point Topology
 Direct Connection between two N_Ports
 No Sharing of Media
 Allows Devices to use Full Bandwidth
 Before the transmission two N_Ports perform a Login to assign the N_Port Address

25
 Arbitrated Loop Topology
 A loop of 127 Ports ( 126 NL_Port , 1FL_Port )
 The Bandwidth is shared by the Active Nodes
 Media Access is gained through an Arbitration Protocol
 Can Connect Ports up to a distance of 10 Km

26
 FC - Switched

Array Array Array Array

Array Array Array Array

Single Switch Fabric

Multi Switch Single Fabric

27
 In a Nutshell
 NAS gateway
– A NAS device without integrated
storage (i.e., just the NAS processor).
– Instead, the NAS device connects
externally to storage by direct
attachment or by a SAN.

 SANergy
– SANergy is software from IBM and
Tivoli that provides NAS-like file sharing,
with data sent over the SAN rather than
the LAN for improved performance.
– IBM NAS gateways include SANergy
function.
28
In a Nutshell . . . NAS

29
In a Nutshell
 NAS : Network Attached Storage
 A NAS device is usually an integrated
processor plus disk storage, is attached to a
TCP/IP-based network (LAN or WAN), and
accessed using specialized file access/file
sharing protocols.
 File requests received by a NAS are
translated by the internal process to device
requests.

30
 In a Nutshell
 The benefits of the major types of processor-to-storage connectivity
– DAS: Optimized for single, isolated processor and low initial cost.
– SAN: Optimized for performance and scalability
• Some of the major potential benefits include support for high-speed Fibre
Channel media which is optimized for storage traffic, managing multiple disk
and tape devices as a shared pool with a single point of control, specialized
backup facilities that can reduce server and LAN utilization and wide
industry support.
– NAS: Optimized for ease-of-management and file sharing using lower-cost
Ethernet-based networks.
• Installation is relatively quick, and storage capacity is automatically assigned
to users on demand

31
 In a Nutshell
 NAS gateways
– Optimized to provide NAS benefits with more flexibility in selecting the disk
storage than offered by a conventional NAS device.
– Gateways can also protect and enhance the value of installed disk systems.
 Tivoli SANergy
– Optimized for data sharing (like a NAS), but at SAN speeds.
– Tivoli SANergy is disk vendor-independent, and can be added to an existing
SAN to enhance its value.

32
 Connectivity
 Processors and storage are physically connected
– Direct attach: A single storage device is connected to a single processor
(host).
– Network attach: One or more processors are connected to one or more
storage devices.
 Media: The type of cabling and associated protocol that provides the connection.
– Ethernet
• A media for building LANs in the 1980s.
• Typical bandwidths are 10Mbps, 100Mbps, and 1Gbps.

33
 Connectivity
– Fibre Channel
• A technology developed in the 1990s that has become increasingly
popular as a storage-to-processor media (for both SANs and DAS).
• Bandwidth is generally 100MBps, with 200MBps expected in 2001.
– Parallel SCSI (Small Computer System Interface)
• An evolving technology with origins in the 1980s.
• Typical bandwidths are 40MBps, 80MBps, and 160MBps.
– SSA (Serial Storage Architecture)
• A media technology optimized for high-performance.
• Bandwidth is 160MBps.

34
Direct Attached Storage (DAS)
 Storage that is generally restricted to
access by a single host (processor);
sometimes by two hosts in small cluster
(failover or failback) configurations.
 The Total Cost of Ownership of DAS may
be higher than for networking approaches
due to the difficulty of sharing unused
capacity with other processors, and the
lack of a central point of management for
multiple disk systems.

35
Storage Area Networks (SAN)
 A dedicated network for storage devices and
the processors that access those devices.
 SANs today usually built using Fibre
Channel technology, but the concept of a
SAN is independent of the underlying type
of networks.
 I/O requests to disk storage on a SAN are
called “block I/Os”, just as for direct-
attached disk.
 The long-term lower Total Cost of
Ownership compared to an alternative
connectivity approach.
36
 Storage Area Networks (SAN) - Benefits
 Access: Longer distance between processors and storage, higher availability,
improved performance (because I/O traffic is offloaded from a LAN to a
dedicated network, and because Fibre Channel is generally faster than most
LAN media).
 Consolidation: Replacement of multiple independent storage devices by fewer
devices that support capacity sharing – this is also called disk and tape pooling.
 Protection: LAN-free backups occur over the SAN rather than the (slower)
LAN, and server-free backups can let disk storage “write itself” directly to tape
without processor overhead.
 Data Sharing: Sharing data offers benefits such as reducing the number of
copies of files, increasing accessibility to current data and reducing the need to
transfer copies of data between servers over the network.
37
Network Attached Storage (NAS)

 A device that resides on a network may be

shared with non-storage traffic.
 Network is usually an Ethernet LAN, but
could be any network that supports the IP-
based protocols that NAS uses.
 In contrast to “block I/O” used by DAS
and SANs, NAS I/O requests are called
“file I/Os”.

38
NAS Gateways

 A NAS gateway provides the function of

a conventional NAS appliance but without
integrated disk storage.
 The disk storage is attached externally to
the gateway, and may also be a standalone
offering for direct or SAN attachment.
 The gateway accepts a file I/O request and
translate that to a SCSI block-I/O request
to access the external attached disk
storage.

39
iSCSI

 A proposed industry-standard that allows

SCSI I/O commands to be sent over a
network using the popular TCP/IP
protocol.
 This is analogous to the way SCSI
commands are already mapped to Fibre
Channel, parallel SCSI, and SSA media.
 Processors supporting iSCSI can attach to
iSCSI storage over a TCP/IP network
such as an Ethernet LAN.

40
 iSCSI
 File sharing
– NAS supports file sharing while iSCSI SANs and Fibre Channel SANs
generally do not.
– The SANergy product can add file sharing capabilities to iSCSI SANs and
Fibre Channel SANs.
 Management
– Backup of data is done through any method that support SCSI-attached
volumes.
– A NAS appliance, because it “hides” disk volumes from its clients and often
includes specialized backup facilities, may be easier to install and manage.
– SANs currently have more storage-related management tools than iSCSI,
such as support for tape sharing for backup.
41
 iSCSI
 Performance
– iSCSI may perform better than NAS due to reduced protocol overhead,
since it handle SCSI directly rather than translating between file-I/O
protocols and SCSI.
– Fibre Channel SANs support commands mapped directly to Fibre Channel
media, and processor overhead for this mapping is low. In SCSI, handling
of the TCP/IP protocol requires processor cycles at both ends.
 Cost
– An iSCSI SAN likely has a lower cost than a Fibre Channel SAN.
– An iSCSI disk device, all else equal, may be lower cost than a NAS
appliance since the iSCSI device does not need to support file systems, file
sharing protocols, and other facilities often integrated into NAS products.

42
 IOPS

• What is IOPS:

• IOPS is a common performance measurement used to benchmark

computer storage devices like hard disk drives (HDD), solid state
drives (SSD), and storage area networks (SAN).

• IOPS measurements commonly seen:

• Total IOPS: Average number of I/O operations per second
• Read IOPS: Average number of read I/O operations per second
• Write IOPS: Average number of write I/O operations per second

43
 Thin disk

• Thin.

A thin LUN lets you overallocate its size. That is, you can assign a LUN
size that is larger than the size of the volume. Even though you
specify the size of a thin LUN when you create it, storage space is
assigned on demand instead of up front. This method greatly
improves the utilization rate of the LUN because storage space is
assigned only as data is written to the LUN. However, the size of the
LUN is reported as the total storage space that you specify when you
create the LUN.

44
 Thick Disk

• Thick.

All storage space that you specify when you create a thick LUN is
allocated up front and the storage space is reserved on the volume.
Snapshots, other LUNs, and shared folders on the volume cannot
consume storage space that is reserved. The size of the LUN is
reported as the total storage space that you specify when you create
the LUN. You cannot assign more storage space than the available
non-reserved storage space on the volume.

45
 Evolution of Storage
Storage Today
Performance/Reliability/Managem

Ethernet Storage Area Network

(SAN)
Network-Attached Storage
(NAS)

Direct-Attach Storage
(DAS)

Internal Storage

Time/Complexity

46