Introduction to Fibre Channel

Copyright © 2007 EMC Corporation. Do not Copy - All Rights Reserved.

Introduction to Fibre Channel - 1

© 2007 EMC Corporation. All rights reserved.

Introduction to Fibre ChannelIntroduction to Fibre Channel

Welcome to Introduction to Fibre Channel.The AUDIO portion of this course is supplemental to the material and is not a replacement for the student notes accompanying this course.EMC recommends downloading the Student Resource Guide from the Supporting Materials tab, and reading the notes in their entirety.

Copyright © 2007 EMC Corporation. All rights reserved. These materials may not be copied without EMC's written consent. Use, copying, and distribution of any EMC software described in this publication requires an applicable software license.THE INFORMATION IN THIS PUBLICATION IS PROVIDED “AS IS”. EMC CORPORATION MAKES NO REPRESENTATIONS OR WARRANTIES OF ANY KIND WITH RESPECT TO THE INFORMATION IN THIS PUBLICATION, AND SPECIFICALLY DISCLAIMS IMPLIED WARRANTIES OF MERCHANTABILITY OR FITNESS FOR A PARTICULAR PURPOSE.Celerra, CLARalert, CLARiiON, Connectrix, Dantz, Documentum, EMC, EMC2, HighRoad, Legato, Navisphere, PowerPath, ResourcePak, SnapView/IP, SRDF, Symmetrix, TimeFinder, VisualSAN, “where information lives” are registered trademarks. Access Logix, AutoAdvice, Automated Resource Manager, AutoSwap, AVALONidm, C-Clip, Celerra Replicator, Centera, CentraStar, CLARevent, CopyCross, CopyPoint, DatabaseXtender, Direct Matrix, Direct Matrix Architecture, EDM, E-Lab, EMC Automated Networked Storage, EMC ControlCenter, EMC Developers Program, EMC OnCourse, EMC Proven, EMC Snap, Enginuity, FarPoint, FLARE, GeoSpan, InfoMover, MirrorView, NetWin, OnAlert, OpenScale, Powerlink, PowerVolume, RepliCare, SafeLine, SAN Architect, SAN Copy, SAN Manager, SDMS, SnapSure, SnapView, StorageScope, SupportMate, SymmAPI, SymmEnabler, Symmetrix DMX, Universal Data Tone, VisualSRM are trademarks of EMC Corporation. All other trademarks used herein are the property of their respective owners.



© 2007 EMC Corporation. All rights reserved. Introduction to Fibre Channel - 2

Course ObjectivesUpon completion of this course, you will be able to:

Identify Fibre Channel:– Architecture– Layers– Topologies

Describe SAN concepts

The objectives for this course are shown here. Please take a moment to read them.




Lesson 1: Fibre Channel OverviewUpon completion of this lesson, you will be able to:

Identify Fibre Channel characteristics and utilities

Describe different Network Storage Technologies using Fibre Channel

Identify management tools in a SAN environment

The objectives for this lesson are shown here. Please take a moment to read them.




Fibre ChannelFibre Channel is a serial data transfer interface intended for connecting high-speed storage devices to computers

The high-speed is obtained through the following processes:– Networking and I/O protocols (such as SCSI commands) are

mapped to Fibre Channel constructs– And encapsulated and transported within

Fibre Channel frames– With this, the high-speed transfer of multiple

protocols over the same physical interface is possible

Fibre Channel is a serial data transfer interface that operates over copper wire and/or optical fiber at data rates up to 4 GB/s (gigabits per second) and up to 10 GB/s when used as ISL (E-ports) on supported switches.

Networking and I/O protocols (such as SCSI commands) are mapped to Fibre Channel constructs, and then encapsulated and transported within Fibre Channel frames. This process allows high-speed transfer of multiple protocols over the same physical interface.

Fibre Channel systems are assembled from familiar types of components: adapters, hubs, switches and storage devices.

Host bus adapters are installed in computers and servers in the same manner as a SCSI host bus adapter or a network interface card (NIC).

Hubs link individual elements together to form a shared bandwidth loop.

Fibre Channel switches provide full bandwidth connections for highly scalable systems without a practical limit to the number of connections supported (16 million addresses are possible).

The word fiber indicates the physical media. The word fibre indicates the Fibre Channel protocol and standards.




ChannelsChannel connections, such as parallel bus and tag, ESCON, and SCSI provide fixed connections between host systems and their peripheral devices

Some characteristics of channel technologies are:– High performance– Low protocol overhead– Static configuration– Short distance (although ESCON is somewhat of an exception)– Connectivity within a single system

The next few slides explain how channels and network technologies help create Fibre Channel technology.

Traditionally, host computer operating systems communicated with storage devices over channel connections, such as parallel bus and tag, ESCON, and SCSI. These channel technologies provide fixed connections between host systems and their peripheral devices.

Static connections are defined to the operating system in advance. Tight integration between the transmission protocol and physical interface minimize the overhead required to establish communication and transport large amounts of data to statically defined devices.

Some characteristics of channel technologies are:High performanceLow protocol overheadStatic configurationShort distance (although ESCON is somewhat of an exception)Connectivity within a single system




NetworkNetwork technologies offer more flexibility and distance capabilities than channel technologies

Some characteristics of network technologies are:– Low performance– High protocol overhead– Dynamic configuration– Long distance– Connectivity among different systems

Network technologies are more flexible than channel technologies, and provide greater distance capabilities. Most networks provide connectivity between client or host systems, and carry a variety of data between the devices. A simple example is a network of desktop PCs within a company.

This type of setup can provide each PC with connectivity to file and print services, server-based applications, and corporate intranets.

These PCs are networked to provide shared bandwidth and the ability to communicate with many different systems. This flexibility results in greater protocol overhead and reduced performance.

Some characteristics of network technologies are:Low performanceHigh protocol overheadDynamic configurationLong distanceConnectivity among different systems




Fibre Channel captures some benefits of channels and networks

Fibre Channel standards define layered communications architecture similar to other networking environments

Fibre Channel: Channels with Network Characteristics

Channels

(benefits)

Networks

(benefits)

Fibre Channel

Fibre Channel captures some benefits of channels and networks. A Fibre Channel fabric is a switched network, and provides a set of generic, low-level services onto which host channel and network architectures can be mapped.

Fibre Channel standards define layered communications architecture similar to other networking environments. Each level of the Fibre Channel protocol stack provides a specific set of functions.




Network Storage Technologies

Typical applications

Type of data

Key requirement

Type of transport

OLTP, data warehousing, ERP

Fibre Channel(FCP, FICON)IP (iSCSI, FCIP, iFCP)

Block

Deterministicperformance

NASNetwork-Attached

Storage

Software and product development, file server consolidation

File

Multi-protocolsharing

IP, Fibre Channel(*MPFS)

CASContent Addressed

Storage

Contentmanagement

Longevity, integrity assurance

IP

Object, fixed content

SANStorage Area

Networks

The Fibre Channel can be used as a type of transport in SAN and NAS solutions.




Storage Area Network ManagementStorage Area Networks (SANs) are networks of host and storage devices that are often connected over Fibre Channel Fabrics

A common method of managing the variety of devices on a SAN is SNMP (Simple Network Management Protocol)

The FibreAlliance is defining the SNMP MIB (Management Information Base) to facilitate SAN management

The Fibre Channel Management Integration (FCMGMT-INT) MIB provides a heterogeneous method of managing multiple devices across a SAN

The networks of host and storage devices (called Storage Area Networks, or SANs) are often connected over Fibre Channel fabrics. A common method to manage the variety of devices on a SAN is SNMP (Simple Network Management Protocol), and is popular because it’s widely supported and can be run out of band (which is advantageous because it does not rely on the Fibre Channel network).

FibreAlliance, an open industry consortium, is defining an SNMP MIB (Management Information Base) to facilitate SAN management. The MIB is a group of parameters (variables) whose values define and describe the status of a network and its components. The Fibre Channel Management Integration (FCMGMT-INT) MIB provides a heterogeneous method to manage multiple devices across a SAN.




Lesson 2: Fibre Channel ConceptsUpon completion of this lesson, you will be able to:

Describe how Fibre Channel standards are defined

Identify the attributes of FC-0, FC-1, and FC-2

Identify the roles of FC-3 and FC-4 in Fibre Channel protocol





LevelsThe Fibre Channel standards defines a Level protocol stack

Defines transmission protocol utilizing the 8B/10B code/decode method, which improves the transmission characteristics and enhances error recovery.

FC-1

Defines physical media including connectors, cables, transmitter and receiver technology. Supports a variety of data rates.FC-0

Defines the rules by which nodes communicate including data framing, frame sequencing, flow control, and class of service.FC-2

Defines a set of services to support advanced functions.FC-3

Defines the Fibre Channel Link Encapsulation (FC-LE)FC-4

Upper level Protocol - SCSI-3, IP, ESCON/FIPS, etc..ULP

Level FunctionFibre Channel

The Fibre Channel standards define a layered protocol:ULP (Upper level Protocol) is not actually part of Fibre ChannelThe FC-4 level of Fibre Channel is designed to hand off to another protocol such as SCSI. Fundamentally, the commands at FC-4 for SCSI allow SCSI initiators and targets to communicate over Fibre ChannelThe FC-3 Defines a set services to support advanced functionsThe FC-2 level serves as the transport mechanism of Fibre Channel. The transported data is transparent to FC-2 and visible to FC-3 and aboveThe FC-1 defines the transmission protocol including serial encoding and decoding, special characters and error control. Information is encoded 8 bits at a time into a 10 bit transmission character and transmitted over the cable The FC-0 is the lowest level and defines the physical link in the system, including the fiber, connectors, optical and electrical components. This level covers a variety of media and the associated drivers and receivers capable of operating at a wide range of speeds




FC-0

The FC-0 level of FC-PH describes theFibre Channel link– Physical Interface– Optical and Electrical Interfaces– Cables, Connectors, etc

Each fiber is attached to a transmitter of a Port at one end and a receiver of another Port at the other end

The FC-0 level of FC-PH describes the Fibre Channel link. The FC-0 level covers a variety of media and the associated drivers and receivers capable of operating at a wide range of speeds. The FC-0 level is designed for maximum flexibility and allows the use of many technologies to meet the widest range of system requirements.

Each fibre is attached to a transmitter of a Port at one end, and a receiver of another Port at the other end. When a Fabric is present in the configuration, a fibre may attach to an N_Port and an F_Port. Patch panels or portions of the active Fabric may function as repeaters, concentrators or fibre converters.




ConnectorsThe LC assembly design incorporates an RJ-style latch in a connector body with half the footprint of a conventional SC

The SC connector is the standard connector for fiber optic cables. It is a push-pull connector

The connectors, cables, and host Bus adapters (HBA) characteristics that are part of the FC-0 level are reviewed next.

The LC assembly design incorporates an RJ-style latch in a connector body, with half the footprint of a conventional SC. High-precision, 1.25 mm ceramic ferrules and engineered composites provide a durable package with consistent, repeatable performance.

The SC connector is the standard connector for fiber optic cables. It is a push-pull connector and is favored over the ST connector (commonly used in patch panels). If the cable is pulled, the tip of the cable in the connector does not move out, which would result in loss of signal quality.




Multi-Mode Cable Multi-mode transmitters send multiple short-wavelength signals through the same fiber

The angle of entry is high and the signals tend to cancel each other out

Multi-mode fiber has a larger diameter (62.5 or 50 microns) core than single-mode fiber

The multi-mode cable is dominant for distances of 500 meters or less. Multi-mode has an inner diameter of 62.5 or 50 microns, and allows light to enter the cable in multiple modes, including straight and at different angles. The multiple light beams tend to lose shape as they move down the cable. This loss of shape is called modal dispersion and limits the distance.




Single-Mode Cable Single-mode transmitters send one longer-wavelength signal down a much thinner-cored fiber

The angle of entry is low (less bouncing) and there is little to muddy the signal, hence the much greater distances

Single mode cable has an inner diameter of 9 microns

The single mode cable is used for long distance cable runs, and is limited only by the power of the laser at the transmitter and the receiver sensitivity. Single mode cable has an inner diameter of 9 microns and is always used with a long wave laser, which limits the effects of modal dispersion. Therefore, with single mode cables, the distance is greatly increased.




HBA (Host Bus Adapter)An HBA is any adapter that allows a computer bus to attach to another bus or channel

A Host Bus Adapter performs many low-level interface functions automatically to minimize the impact on host processor performance

The HBA enables a range of high-availability and storage management capabilities– Load balance– Fail-over– SAN administration– Storage management

An HBA is an I/O adapter that sits between the host computer's bus and Fibre Channel loop and manages the transfer of information between the two channels. In order to minimize the impact on the host processor performance, the host bus adapter performs many low-level interface functions automatically or with minimal processor involvement.

In simple terms, a host bus adapter (HBA) provides I/O processing and physical connectivity between a server and storage. The storage may be attached using a variety of direct attached or storage networking technologies, including Fibre Channel, iSCSI, VI/IP, FICON, or SCSI. Host bus adapters provide critical server CPU off-load that frees servers to perform application processing. As the only part of a storage area network that resides in a server, HBAs also provide a critical link between the SAN and operating system and application software. In this role, the HBA enables a range of high-availability and storage management capabilities, including load balancing, fail-over, SAN administration, and storage management.




FC-1

Defines the method used to Encode data prior to transmission and decode the data upon reception

A benefit of using 8b/10b encoding is that it defines a number of Special Characters

Whenever the encoder creates a character with more ones than zeros, it remembers this by setting a single-bit variable, called the Current Running Disparity (CRD), to positive. Whenever the encoder creates a character with more zeros than ones, it sets the CRD to negative.

The CRD is fed back to the encoder to select the appropriate encoding in order to balance the number of ones and zero bits transmitted.

As a series of characters are processed, the output alternates between positive and negative disparity.




Encoding ProcessThe encoding process transforms 8-bit input characters into 10-bit transmission characters having the desired attributes

The encoding process transforms 8-bit input characters into 10-bit transmission characters having the desired attributes. Not all of the possible 10-bit patterns are used. 10 bits allow 1024 different combinations.

To prevent excessive DC (direct current) components and run length problems, only characters containing 6 ones and 4 zeros, 5 ones and 5 zeros, 4 ones and 6 zeros are used. Any other weighting of bits is invalid.

5 ones and 5 zeros are considered to have neutral disparity6 ones and 4 zeros are considered to have positive disparity4 ones and 6 zeros are considered to have negative disparityAll possible 8-bit characters have 2 possible 10-bit encodings

This is the same method used to transmit a data stream in ESCON. The purpose is to break up the data stream in the serial environment and, at the same time, allow control characters to be embedded into the steam to speed-up communications.




Ordered SetsTwo types of Transmission Characters (Data and Special) are defined

Fill Bytes

Payload (information being

Transported

Optional Header

Optional Header

Data Field

Header

IDLE

Certain combinations of Transmission Characters, referred to as Ordered Sets, are designated to have special meaning. Ordered Sets are used to identify frame boundaries, transmit primitive function requests, and maintain proper link transmission characteristics during periods of inactivity.




FC-2Defines the structure of the Fibre Channelframe

The transported data is transparent to FC-2 and visible to FC-3 and above

The FC-2 level serves as the transport mechanism of the Fibre Channel. The transported data is transparent to FC-2 and visible to FC-3 and above.




FramesFrames are the basic building blocks of a Fibre Channel connection

All information in Fibre Channel is passed in frames

Fill BytesPayload (information being Transported

Optional Header

Optional Header

Data Field

Header

Star

t-of-F

ram

e

CR

C

End-

of-F

ram

e

Frames are the basic building blocks of a Fibre Channel connection. The frames contain the information to transmit, the address of the source and destination ports, and link control information.

All information in Fibre Channel is passed in frames. The maximum amount of data carried in a frame is 2112 bytes; the total frame size is 2148 bytes. The general structure of a Frame is specific.




Header ContentsThe header contains the Source and Destination Addresses which allows the frame to be routed to the correct port

Fill BytesPayload (information being Transported

Optional Header

Optional Header

Data Field

Header

Star

t-of-F

ram

e

CR

C

End-

of-F

ram

e

Parameter Field (PARM)

OX_IDOX.ID

DF_CTLDF_CTLSEQ_ID

Frame Control (F_CNT)TYPE

Source Address (S-ID)CS_CTL

Destination Address (D-1D)R_CTL

The header contains the Source and Destination Addresses, which allow the frame to be routed to the correct port. The Type field interpretation is dependent on whether the frame is a link control or Fibre Channel data frame. For example, if the frame is a data frame, a 08 in the type field indicates SCSI FCP information in the Data field.




Exchange and SequencesAn Exchange is a unidirectional or bi-directional set of non-concurrent Sequences

FC-2 manages Exchanges that map directly to operations

A Sequence is contained within an Exchange and is comprised of one or more Frames

The purpose of the Sequence is to reorder data when it is received at the other end

An Exchange is a uni or bi-directional set of non-concurrent Sequences. An Exchange is the largest construct understood by FC-2. FC-2 manages Exchanges that map directly to operations.

FC-2 manages Sequences as unidirectional transfers of one or more frames. A Sequence is contained within an Exchange and is comprised of one or more Frames. FC-2 names each Sequence and tracks each Sequence to completion. The purpose of the Sequence is to reorder data when it is received at the other end.




R_RDY

Buffer to Buffer CreditR_RDY is sent to a transmitting node as buffers are cleared

An FC switch also performs buffering and flow control internally and externally

Data Frame Data Frame

R_RDY

N_Port N_PortSwitch

The FC-2 provides flow control for buffer management. When nodes initialize on the fabric, they agree on operational parameters such as the number of buffers available (Buffer Credits). Transmitting nodes can continue to transmit as long as there are buffer credits. R_RDYs are sent to a transmitting node as buffers are cleared. An FC switch also performs buffering and flow control internally and externally.




Classes of Service

Buffer-to-Buffer flow controlFabric can reject framesFabric busies framesDelivery Order Guaranteed

OptionalNx_Port supports Class of serviceOptionalFabric support Class of service

End-to-End flow controlOptionalUnidirectional Connection

OptionalOptionalBuffered ClassOptionalOptionalCamp-onOptionalOptionalDedicated SimplexOptionalOptionalStacked connect Request

OptionalBroadcast supportOptionalMulticast supportOptionalHunt group support

Initial roundtrip delay1:11 to Many1:1Communications type between ports

Fabric discards FramesClass 5Class 4 Class 3 Class 2Class 1Functions

Classes of Service

Classes of Service are different types of topology independent services provided by the Fabric and used by the communicating N_Ports destination. The allocation and retention method between the N_Ports and level of delivery integrity required for an application distinguish classes of service. If the Fabric is not present, the service is provided as a special case of point-to-point. Fabrics and N_Ports are not required to support all Classes of service.

The Classes of Service table is shown here. Please take a moment to review.




Fibre Channel addressingFibre Channel Addresses are required to route the frames from source to target

24 bits (3 bytes) physical addresses are assigned when a Fibre Channel node is connected to the switch (or loop in the case of FC-AL)

Source

Target

FC Initiator:

HBA

FC Responder:

Symmetrix FA

Or

CLARiiON SP Ports

FC Switch

Fibre Channel addresses are used to designate the source and destination of frames in the Fibre Channel network. The Fibre Channel address field is 24 bits/3 bytes in length. Unlike Ethernet, these addresses are not burned in, but are assigned when the node enters the loop or is connected to the switch. There are reserved addresses, which are used for services rather than interface addresses.




Addressing LayoutPhysical Address is switch specific and dynamically generated in the Fabric Login (FLOGI)

Each N_Port has a Fabric-unique identifier, with the following layout

1FPort in Switch

0002AL_PASwitch

FC-SW FC-AL

The Physical Address is switch-specific and dynamically generated in the Fabric Login (FLOGI).

The Fabric Logon is discussed later in this lesson.

Address identifiers are three bytes in length. The Frame Header contains two three-byte fields for address identifiers, the D_ID and S_ID fields. Each N_Port has a fabric-unique identifier, the N_Port Identifier, by which it is known. The source and destination N_Port Identifiers and alias address identifiers are used to route frames within the fabric.




Addressing layout1. Domain ID: Identifies the source or target switch inside

the Fabric2. Switch Port: Identifies the source or target port in the

switch3. AL_PA - Used in Private Loop environment, identifying

the NL_Port (node loop port)

FC-SW FC-AL

02Switch

1FSwitch

1FPort in Switch

12AL_PA

12AL_PA

02Switch

1 2 3

The most significant 8 bits of the Fibre Channel address contain the Domain ID, which basically identifies the switch.

In the Fabric environment, this allows frames to be routed between switches.

The middle 8 bits contain the area address, which is implemented as the port address within the switch.

In the Fabric environment, this allows frames to be routed between switches.

In the Private Loop environment, the Domain and Area fields contain zeros, and the port field contains the AL_PA (Arbitrated Loop Physical Address) for the NL_Port.

Note:

With a McDATA switch, the third byte is always 13.




World Wide NameA World Wide Name, or WWN, is a 64-bit address used in Fibre Channel networks to uniquely identify each element in the network

Assigned to a host bus adapter (HBA) or switch port by the vendor at the time of manufacture, it is similar to the MAC address in an Ethernet network

1 0 0 0 0 0 0 0 C 9 2 0 D C 4 0Example: Emulex HBA’s World Wide Name

Reserved

12 bits

Company OUI

23 bits

Vendor Assigned

24 bits

A World Wide Name, or WWN, is a 64-bit address used in Fibre Channel networks to uniquely identify each element in the network.

Assigned to a host bus adapter (HBA) or switch port by the vendor at the time of manufacture, it is similar to the MAC address in an Ethernet network.

There are two designations of WWN; World Wide Port Name and World Wide Node Name. Both are globally unique 64-bit identifiers. The difference lies in where each value is “physically” assigned. For example, a server may have dual HBAs installed, thus having multiple ports or connections to a SAN. A WWPN is assigned to each physical port.

The WWNN represents the entire server, which can be referred to as the node or node process, and is derived from one of the WWPNs. EMC uses the WWPN for all configurations

Fibre Channel specifications allow for multiple formats of the World Wide Name. The example shown is that of the IEEE Registered Name Format.

NAA is the Name Assignment Authority, which assigns the 24 bits (IEEE Company ID) to the specific vendor (i.e. EMC).

Values for World Wide Name formats are based on the IEEE company_id. More information on these formats can be found at www.standards.ieee.org.




World Wide Name – Target (FA and SP ports)Symmetrix and CLARiiON ports are soft-assigned based on slot number, processor, port and array serial number

5 0 0 6 0 4 8 2 E 8 9 1 2 B 9 0Example: Symmetrix FA World Wide Name

5 0 0 6 0 1 6 0 0 0 6 0 0 1 B 2Example: CLARiiON SP World Wide Name

The WWN for Symmetrix FAs and Clariion SPs ports are soft-assigned, so they remain unchanged after a component replacement.




World Wide Name – Symmetrix TargetSymmetrix 8000 and previous

15bA195183790

011110001011101000100100010010101110Symmetrix Serial Number 30 bits FA 4 bitsSidePortEMC Company ID 24 bits

10 11101010110010000110011000111000101000010000000110000000000101

5 0 0 6 0 4 8 2 E 8 9 1 2 B 9 0

Symmetrix DMX

Symmetrix Serial Number 29 bits010

2CB187900106

10000101011001100110010000011001010Half FA 4 bitsSidePortEMC Company ID 24 bits

01 00011000100011100011001100110011000010000000110000000000101

5 0 0 6 0 4 8 A C C C 8 3 2 A 1

Because the WWN of the FA is dependent on the Symmetrix Serial Number and the Slot rather than being “burned in”, the WWN stays constant if the FA fails and must be replaced. To determine the WWN of an FA depends on the version of the Symmetrix being used. The original calculation was that the last 6 bits of the WWN were used. Bit position 6 specified the port (0 or 1), Bit position 5 the processor ( 0-A/1-B ) and Bit position 1-4 the slot (add 1 to get the adapter).

On the DMX series, the ports are laid out as follows, from the top down:DB - processor 4, port 1DA - processor 4, port 0CB - processor 3, port 1CA - processor 3, port 0BB - processor 2, port 1BA - processor 2, port 0AB - processor 1, port 1AA - processor 1, port 0

The new calculation had to be both backward compatible with previous Symmetrix versions, and at the same time allow for more addresses. The new calculation borrows a bit from the serial number. It was determined that we do not need 30 bits for a serial number, 29 gives us all the range we need. So, bit position 30 is now dubbed a "half bit" The half bit determines which pair of processor we are working with. It is set to 0 for processors A & B, and 1 for C & D. Then, as above, we make use of the “processor" bit. This is used along with the half bit to determine exactly what processor. It breaks down like this:Half bit 0 / Processor bit 0 – AHalf bit 0 / Processor bit 1 – BHalf bit 1 / Processor bit 0 – CHalf bit 1 / Processor bit 1 - D

Then, like the original calculation, we use the Port bit to determine the 0/A port or the 1/B port.




World Wide Name – CLARiiON TargetCX600 example:

CLARiiON seed 32 bits

2 5PortEMC Company ID 24 bits

01010010 0000101100010000000001100000000000000110000100000110000000000101

5 0 0 6 0 1 6 0 0 0 6 0 0 1 B 0

CLARiiON - CX600WWN seed - 00:60:01:b2Therefore, the resulting WWNN is 50:06:01:60:80:60:01:b2The resulting WWPNs are:Storage Processor APort 0 - 50:06:01:60:00:60:01:b2Port 1 - 50:06:01:61:00:60:01:b2Port 2 - 50:06:01:62:00:60:01:b2Port 3 - 50:06:01:63:00:60:01:b2

Note:

On the FC4500, we have the only real exception as the FC ports are in fact connected to a hub and share their WWPN. To get around this, only one port from each SP would be connected to the Fabric. Two ports existed per SP to allow for a dual cluster direct attach.




Frames Routing

Source

TargetFabric

Physical cable connected

HBA WWN:

10000000C920DC40

FA WWN:

50060482E8912B90


ISL

Sw 01

Sw 02

Port 07

Port 05

When N_Ports are connected to F_Ports, the Fabric Login begins, associating physical address and World Wide Name

When N_Ports are connected to F_Ports, the Fabric Login begins associating the physical address and worldwide name.




Frames Routing (Cont.)

Source

TargetFabric


HBA WWN:

10000000C920DC40

FA WWN:

50060482E8912B90


ISL

Sw 01

Sw 02

Port 07

Port 05

The WWN and Physical Address association is done by the switch and is stored using internal tablesSo the frames are routed using their Physical Address

WWN -> Switch/Port/AL_PA

10000000C920DC40 -> 020500

50060482E8912B90 -> 010700

The WWN and Physical Address association is done by the switch and stored using internal tables.

The frames are routed using their Physical Address, as happens in an Arbitrated Loop environment using just the Arbitrated Loop Physical Address.




Frames Routing (Cont.)

Source

TargetFabric


HBA WWN:

10000000C920DC40

FA WWN:

50060482E8912B90


ISL

Sw 01

Sw 02

Port 07

Port 05

Any changes in the switch’s ports do not impact future frame routing, once internal tables are updated

WWN -> Switch/Port/AL_PA

10000000C920DC40 -> 020600

50060482E8912B90 -> 010700Port 06

When nodes generate frames, they are routed by their addresses, not their WWNs. However, tables can be built which can associate WWNs to the destination addresses. A World Wide Name is a 64 bit value (16 characters).




Fibre Channel Logins

N_Port 1

N_Port 1

F_PortF_Port F_PortF_Port N_Port 2

N_Port 2

Fabric

Process aProcess a

Process bProcess b

Process cProcess c

Process xProcess x

Process yProcess y

Process zProcess z

Fabric Login Fabric LoginAccept Accept

Port LoginAccept

Process LoginAccept

There are three types of “login” supported in Fibre Channel: Fabric, Port and Process. All node ports must attempt to log in with the Fabric. This is typically done right after the link or the Loop is initialized. It consists of the node port transmitting a Fabric Login (FLOGI) frame to the well-known Fabric address hex'FFFFFE'. The normal response is an Accept (ACC) frame from the Fabric back to the node port. Fabric Login accomplishes the following things:

It determines the presence or absence of a Fabric. If a Fabric is present, it provides a specific set of operating characteristics associated with the entire Fabric, including which Classes of service are supported. If a Fabric is present, it will optionally assign or will confirm the native N_Port Identifier of the N_Port that initiated the Login. If a Fabric is present, it initializes the buffer-to-buffer credit.

Before a node port can communicate with another node port, it must first perform N_Port Login with that node port. Similar to Fabric Login, the N_Port transmits a PLOGI frame to the destination node port. Again, the normal response is an ACC frame. Port Login accomplishes the following things:

It provides a specific set of operating characteristics associated with the destination N_Port, including Class of service.With Class 3 services, buffer-to-buffer credit is initialized.

PRLI is an acronym for Process login. Process logins establish sessions between related processes on a source and target N_Port. The processes are typically FC-4 layer applications.




Fabric LoginFabric Login (FLOGI) is used by an N_Port to determine the presence of a Fabric, then exchange service parameters

N_Port performs a login to address FFFFFE (F_Port Server) using a source address of 000000

The Fabric Login service returns a frame which assigns an address to the port (24 bits)

When a Fibre Channel device is attached to a fabric, it begins a fabric login (FLOGI). FLOGI is an extended link service command that sets up a session between two participants. With FLOGI, a session is created between an N_Port or NL_Port and an F_Port or FL_Port. An N_Port sends a FLOGI frame that contains its Node Name, N_Port Name, and service parameters to a well-known address.




FLOGI

Port Address RequestPort Address Request

Registration Request Registration Request

Directory Server (NameServer)

(x’FF FF FC’)Port Identifier (SID)

Port Name (WWN)

Classes of Service (Class 3)

FC-4 Types Supported (SCSI)

Port Type (N_Port)

Zoning Information

HBAN_Port

HBAN_Port

FAN_Port

FAN_Port

Fabric Server

(x’FF FF FE’)

Query Request Query Request

When the N_Port logs in, it uses a 24-bit port address of 0. Because of this, the fabric is allowed toassign the appropriate port address to that device, based on the Domain-Area-Port address format. The newly assigned address is contained in the ACC response frame.




Fabric Login in Loop EnvironmentA public loop port first opens the destination AL_PA 0x00 before issuing the FLOGI

NL_Port logs in a similar process, except that the last 8 bits of the Physical Address is used

Before an NL_Port logs in, it goes through the LIP; this ensures that the switch assigned AL_PA does not conflict with any previously selected AL_PAs on the loop

An NL_Port first opens the destination AL_PA 0 before issuing the FLOGI request. In both cases, the switch accepts the login and returns an accept (ACC) frame to the sender. If some of the service parameters requested by the N_Port or NL_Port are not supported, the switch sets the appropriate bits in the ACC frame to indicate this.

When the NL_Port logs in, a similar process starts, except that the least significant byte is used to assign AL_PA and the upper two bytes constitute a fabric loop identifier. Before an NL_Port logs in, it goes through the LIP on the loop, which is started by the FL_Port. From this process, an AL_PA has already been derived. The switch then decides if it will accept this AL_PA or not. If not, a new AL_PA is assigned to the NL_Port, which then causes the start of another LIP. This ensures that the switch assigned AL_PA does not conflict with any previously selected AL_PAs on the loop.




Fabric Login Name and Directory ServiceInformation is then registered with the Directory Service (FFFFFC) such as:– Port Identifier (ID) = S_ID– Port Name(PN) = WWN of the N_Port– Classes of Service (CS) = Class 3 currently– FC-4 Types Supported (FT) = SCSI-3– Port Type (PT) = N_Port

Once a port has logged in, it can query the Name Service database for information about all other logged in ports

After the node gets its Fabric address from FLOGI, it must register with the SNS with port login (PLOGI) on address 0xFFFFFC. The device may register values for all or just some database objects, but the most useful are:

Port Identifier (ID) = S_IDPort Name(PN) = WWN of the N_PortClasses of Service (CS) = Class 3 currentlyFC-4 Types Supported (FT) = SCSI-3Port Type (PT) = N_Port

Once a port has logged in, it can query the Name Service database for information about all other logged in ports.




Port Login (PLOGI)Happens when an N_Port logs in to another N_Port (HBA logs in to an FA for instance). At this moment, a table is created speeding up data transfer between the nodes

Port logins exchange information such as:– Host address (SID)– Frame size (receive buffer size)– Flow control & version information (TOVs)– Port name (WWPN)

Some platforms require re-login when there is a long idle period (HP-UX for instance)

Service parameters are exchanged between nodes before any upper level commands can be issued

When an N_Port logs in to another N_Port, a table can be built which will keep track of the WWN of the logged in port along with its Fibre Channel address. For example, when the NT and SUNs login to the Symmetrix, a table is created. This speeds data transfers and node-to-node communications.

Some Host types (for example, HP-UX) may not maintain a connection while idle and would need to re-login.

Port login is used to establish a session between two nodes, swapping service parameters and making themselves known to each other.

Port logins exchange information such as:host address (SID)frame size (receive buffer size)flow control & version information (TOVs)port name (WWPN)




Process Login (PRLI)Process Login sets up the environment between node ports and is done by the FC-4 layer

At this level, each Fibre Exchange is composed of SCSI tasks (individual or grouped commands)

Some SCSI commands:– FCP_CMND– FCP_XFER_RDY– FCP_DATA– FCP_RSP

Process login is used to set up the environment between related processes on node ports.

The Fibre Channel Protocol for SCSI-3 standard specifies that each SCSI task corresponds to Fibre Channel exchange. A Fibre Channel exchange consists of a single SCSI command or a group of linked SCSI commands. The FCP mapping of SCSI-3 to Fibre Channel defines four information sets that are transferred between SCSI initiator and target. The information sets are modeled after the SCSI-3 architecture defined protocols services.

FCP_CMND: Corresponds to Send SCSI Command protocol service.FCP_XFER_RDY: Transports the offset and request byte count objects of the Send Data-In and Receive Data-Out protocol services.FCP_DATA: Transports the data object of the Send Data-In and Receive Data-Out protocol services.FCP_RSP: Corresponds to the Send Command Complete protocol service.




FC-3FC-3 was put into Fibre Channel as a placeholder

In concept, FC-4 would pass requests to FC-3 that would perform the desired service and then pass onto FC-2

The FC-3 was put into Fibre Channel as a placeholder. In concept, FC-4 would pass requests to FC-3 that would perform the desired service and then pass onto FC-2.

Some things that have been identified as probably fitting into FC-3 are:Data StripingMultipathingMirroringRAIDData encryptionData compressionData Translation




FC-4The FC-4 level of Fibre Channel is designed to hand off to another protocol, such as SCSI

Fundamentally, the commands at FC-4 for SCSI allow the SCSI initiator and target to communicate over Fibre Channel

The FC-4 level consists of several standards documents that describe how different upper-level protocols use the transport services provided by levels FC-2, FC-1 and FC-0. The purpose of an FC-4 protocol mapping is to make a logical connection between the ULPs and Fibre Channel's transport facilities.




ULP - Upper Layer ProtocolULP not actually part of Fibre Channel

Some examples of ULPs are:– Small Computer System Interface (SCSI)– Intelligent Peripheral Interface (IPI)– High Performance Parallel Interface (HiPPI)– Bus and Tag (FIPS) or ESCON– IEEE 802.2 Logical Link Control (LLC)

FC-0 Physical LayerFC-1 Encode and DecodeFC-2 Routing, Flow Control

Common ServicesFC-3FC-4 Mapping InterfaceULP Upper Layer Protocol

ULP (Upper level Protocol) is not actually part of Fibre Channel.

Some examples of ULPs are:Small Computer System Interface (SCSI)Intelligent Peripheral Interface (IPI)High Performance Parallel Interface (HiPPI)Bus and Tag (FIPS) or ESCONIEEE 802.2 Logical Link Control (LLC)




Lesson 3: SAN ConceptsUpon completion of this lesson, you will be able to:

Describe Direct Connect, Arbitrated Loop and Switched Fabric topology

Identify the differences between a Hub and Switch

Identify: – Fibre Channel Port types– Departmental Switches (models and features)– Enterprise Directors (models and features)

Describe different SAN usages (VSAN, Routed and Virtualization)





Physical and Logical TopologiesThe Fibre Channel environment consists of physical topology and logical topology

Fibre Channel Switch

WindowsServer

SunServer

Storage

PhysicalTopology

LogicalTopology

PhysicalTopology

The Fibre Channel environment consists of a physical and logical topology. The physical topology describes the physical interconnects among devices like servers, storage, and switches. The logical topology describes the logical paths established between the operating system device names and their associated storage ports and volumes.

Logical topologies in the EMC/Fibre Channel switch environment can generally be described in terms of fan-in (into the EMC storage array) and fan-out (out of the EMC storage array).




Direct Connect TopologyThe ANSI Fibre Channel Standards define three topologies:– Direct Connect– Arbitrated loop (FC-AL)– Switched Fabric (FC-SW)

Direct Connect is the simplest topology, where two devices are directly connected to each other

The ANSI Fibre Channel Standards define three topologies: Direct Connect, Arbitrated loop, and Switched fabric.

This slide displays a Fibre Channel topology where two devices are directly connected to each other.

EMC specifies a 2 node Fibre Channel Arbitrated Loop connection called Direct Connect. This Topology is the same as the standard Arbitrated Loop, except there are two nodes in the Loop and the Arbitrated Loop Switch is not used.




Arbitrated Loop (FC-AL) TopologyAll devices are in a loop or ring over attachment points called L_Ports (loop ports)

FC-AL is a low-cost connectivity solution because it does not require switches

Fibre Channel Hub

Efficiency and connectivity is enhanced by incorporating one or more hubs into the loop

Arbitrated Loop (FC-AL) is a daisy-chain connecting up to 126 devices in a loop configuration over attachment points called L_Ports (loop ports). FC-AL is a low-cost connectivity solution because it does not require switches.

FC-AL is a good choice for small to medium-sized configurations, and provides a growth path by allowing connection of a loop to a switched Fabric.

Efficiency and connectivity is enhanced by incorporating one or more hubs into the loop. Routing traffic through a hub on each leg of a loop eliminates the loss of the entire loop, as happens in a hubless loop.




Switched Fabric (FC-SW) TopologySwitched Fabric is a Fibre Channel topology where many devices connect with each other via Fibre Channel switches

Frames are routed between source and destination by the Fabric

Fibre Channel Switch

This topology allow the most number of connectivity with a theoretical 16 million devices per Fabric

A Switched Fabric is one or more Fibre Channel switches that connect multiple devices. Rather than travel around an entire loop, frames are routed between source and destination by the fabric. This topology allows the most number of connectivity, with a theoretical 16 million devices per fabric.




Arbitrated Loop SwitchesFibre Channel hubs are used with Fibre Channel Arbitrated Loop (FC-AL) to increase server and storage connectivity

Using a hub, multiple servers can access multiple storage devices

EMC Symmetrix storage systems are qualified with hubs in limited configurations with:– HP-UX – Sun Solaris– Windows NT, Windows 2000, Windows

2003 – Siemens servers

Fibre Channel hubs are used with Fibre Channel Arbitrated Loops to increase server and storage connectivity. This way, multiple servers can access multiple storage devices simultaneously.

A hub can only have one Full Duplex connection between port pairs at any one time. Although every port is capable of full speed operation, only one pair of ports can be active at any one time. This is because only two nodes in a FC-AL environment can be active at any one time and they must communicate with each other.

EMC Symmetrix storage systems are qualified with hubs in limited configurations with HP-UX, Sun Solaris, Windows NT, Windows 2000, Windows 2003, and Siemens servers.




SwitchesA Fibre Channel switch is a device that routes data between host bus adapters and fibre adapters on storage systems

A Fabric is a single switch or multiple switches that interconnect various N_Ports, and is capable of routing frames by using only the D_ID information in a Fibre Channel frame header

Larger Fabric are built by connecting multiple switches through E_Ports

Switched Fabrics offer strong advantages over hubs, in both performance and manageability. For example, a hub can only have one full-duplex connection between port pairs at one time. Although every port in a 10-port hub is capable of full-speed operation, only one pair of ports can be active at any one time. A switched Fabric, however, can have every port pair running full duplex simultaneously at 200 megabytes per second. In a switched topology, aggregate bandwidth of 200 megabytes per second times the Fabric port count is possible.

The aggregate bandwidth of a hub is fixed, but the aggregate bandwidth of a Fabric scales as additional port pairs are added. This scaling capability is an important management feature, because it allows the administrator of the enterprise storage network to reallocate the bandwidth of the network among N_Ports without moving cables and reconfiguring the environment.

A second management benefit of switched Fabrics over hubs is a Fabric-based service called the name server, which maintains a table of all logged-in devices.

Used by the N_Ports for device discovery, the table is maintained by the Name Server during Fabric reconfigurations.




Fibre Channel SAN Switches and DirectorsDirectors are considered to be more highly available than switches

Switches are Directors are

Lower number of ports

Availability features

High performance

Web-based management features

Highest port count

Highest availability

Highest performance

Web-based and/or console-based management features

In a SAN, implementation is necessary to select the director, switch or hybrid technology solution.

The switch/director decision is often financial. Technically, directors are preferable to switches. If we consider the differences between them:

Directors are easier to manage because the relationship between ports is equal and, therefore, it does not need to be considered in performance planning. A single high-availability Fabric is possible using directors, though switches require two Fabrics.Directors have scaled failure localization. This means the GBIC, power supply, etc. can each be replaced without bringing down the entire device, limiting risk of catastrophic failure.Switches offer some localization, at the GBIC or power supply, but in general, a physical failure results in whole device replacement.

Directors are considered to be more highly available than switches. To be fully redundant, there is a requirement for multiple connections to more than one network, with complete routes to all devices throughout those networks. This improves resiliency of the Fabric, allowing the SAN to maintain its functionality in the even of failures. Because fault isolation in a director is at a more granular component level and the components are already dual redundant, directors have the least adverse impact on availability.




EMC Connectrix Products

SERVICE LEVEL

CO

ST

ED-140M

DS-4400M

MDS-9120

MDS-9140

DS-220B

ED-48000B

MDS-9509-V2

Connectrix Directors“Redundant everything” provides optimal serviceability and highest availabilityData-center deploymentMaximum scalabilityLarge Fabrics

Connectrix SwitchesRedundant fans and power suppliesHigh availability through redundant deploymentDepartmental and data-center deployment

DS-4700M

ED-10000M

MP-1620M

MDS-9506-V2MDS-9216i/A

AP-7420B

MP-2640M

DS-4100B

MDS-9513

MDS-9020

DS-4900B

Only EMC offers a complete range of SAN products—from Connectrix directors for data-center deployments to Connectrix switches for data-center and departmental deployment.

The Connectrix Family gives customers:The widest range of SAN service levelsMore choice Additional Fibre Channel functionality A pathway to IP SANsAn additional platform for future network-hosted applications




Types of Fibre Channel Port

Universal port, Connectrix B series equivalent to a G_PortU_Port

General port, a Connectix McDATA switch port with the ability to function as either an F_Port or an E_portG_Port

Expansion port on a switch. Links multiple switchesE_Port

Fabric Loop port, a Fabric port which connects to a NL_PortFL_Port

Fabric port, the access point of the Fabric which connects to a N_PortF_Port

Node Loop port, a port which supports the arbitrated loop topologyNL_Port

Node port, a port at the end of a point-to-point linkN_PortDESCRIPTIONTYPE

In an environment using Host Bus Adapters and Symmetrix FC Directors, ports are configured as either N-Ports or NL-Ports:

N_Port: Node port, a port at the end of a point-to-point linkNL_Port: Node Loop port, a port which supports the arbitrated loop topology

Fibre Channel Switch ports are also configured for specific applications:F_Port: Fabric port, the access point of the Fabric which connects to a N_PortFL_Port: Fabric Loop port, a Fabric port which connects to a NL_PortE_Port: Expansion port on a switch. Links multiple switchesG_Port: General port, a Connectix McDATA switch port with the ability to function as either an F_Port or an E_portU_Port: Universal port, Connectrix B series equivalent to a G_Port




Fabric

N_PortN_Port

N_PortN_Port

N_PortN_Port

N_PortN_Port

F_PortF_Port

F_PortF_Port

F_PortF_Port

G_PortG_Port

E_PortE_Port

F_PortF_Port

E_PortE_Port

Fabric

ISL

A Fabric is a switch or a group of switches linked together.

When looking at the graphic above you will notice that the “cloud” of the Fabric is actually two separate switches attached together via an ISL (Interswitch Link). Attached to the Fabric are several N ports which represent end port devices such as HBAs and Storage ports.

The term Enterprise Director indicates a switch that has all major components redundant at the hardware level. If any major part fails, the switch will automatically fail over, maintaining operation during the failure. The Enterprise directors are commonly used as core switches in a core edge topology.

A Departmental switch has less redundancy and is meant for smaller workgroups. With the lack of redundancy, the departmental switches are commonly used as edge switches for core edge designs with PowerPath in place for redundancy in case of a switch failure.




ZoningPartitions a Fibre Channel switched Fabric into subsets of logical devices

Zones contain a set of members that are permitted to access each other

A member can be identified by its Source ID (SID), its World Wide Name (WWN), or a combination of both

Zoning is used to partition a Fibre Channel switched Fabric into subsets of logical devices. Each zone contains a set of members that are permitted to access each other. Members can be switch ports, HBAs, or storage ports. When zoning is enabled, members in the same zone can see and communicate with each other, but members in separate zones cannot. Ports and devices distributed across multiple switches in a Fabric may be grouped into the same zone.

A zone member can be identified by its Source ID (SID), its World Wide Name (WWN), or a combination of both. Zones can be created using each of these forms of member identification. Thus, there are essentially three types of zoning: hard zoning, soft zoning, and mixed zoning.




What is Zoning?

10000000C920C4E4

10000000C920C321

200000E069000CFE

200000E0690014B8

200000E06900156D

50060482B8912B9E

50060482B8912B8E

50060482B8912B8F

50060482B8912B9F

500601684003491C

A Switched Fabric can be subdivided into a number of zones:A single zone typically includes two or more portsZones can be created by− Switch Port Number (Static)−HBA WWN (Flexible)−Custom Nickname/Alias

EMC recommends that zones include only one HBA (Single HBA zoning), however an HBA may belong to multiple zones. Zones may include one or more EMC Symmetrix FA ports. FAs may belong to multiple zones.




Single Initiator ZoningAlways put ONLY one HBA in a zone with Storage ports

Each HBA port can only “talk” to Storage ports in the same zone

HBAs & Storage Ports may be members of more than one zone

HBA ports are isolated from each other to avoid potential problems associated with the SCSI discovery process

Decreases the impact of a Registered State Change in a Fabric by reducing the amount of nodes that must login again

Under single-HBA zoning, each HBA is configured with its own zone. The members of the zone consist of the HBA and one or more storage ports, such as a Symmetrix Fibre Adapter port, with volumes the HBA will use.

This zoning practice provides a fast, efficient, and reliable means to control the HBA discovery/login process. Without zoning, the HBA attempts to log in to all ports on the Fabric during discovery and during the HBA’s response to a state change notification. With single-HBA zoning, the time and Fibre Channel bandwidth required to process discovery and the state change notification are minimized.

Here are two very good reasons for Single HBA Zoning: First, it cuts down on the reset time for any change made in the state of the Fabric. And secondly, only the nodes within the same zone are forced to log back into the Fabric after a Registered State Change Notification.

When a node’s state has changed in a Fabric, it has to perform the Fabric Login process again before resuming normal communication with the other nodes it is zoned with. If there is only one SCSI Initiator in the zone, then the amount of disrupted communication is reduced.

If a zone has two HBAs and one of them had a state change, then BOTH are forced to log in again, causing disruption to the other HBA that did not have any change in its Fabric state. Performance can be severely impacted by this.




Hard and Soft Zoning

Host

WWPN = 10:00:00:00:C9:20:DC:40

WWPN = 10:00:00:60:69:40:8E:41Domain ID = 21Port = 1

FC Switch

FC Switch Storage

Fabric

WWPN = 50:06:04:82:E8:91:2B:9E

WWPN = 10:00:00:60:69:40:DD:A1Domain ID = 25Port = 3

In general, zoning can be divided into two categories: Port zoning and WWN zoning.




WWN Zoning

Host

WWPN = 10:00:00:00:C9:20:DC:40


FC Switch

FC Switch Storage

Fabric

WWPN = 50:06:04:82:E8:91:2B:9E


WWN Zone 1 = 10:00:00:00:C9:20:DC:40; 50:06:04:82:E8:91:2B:9E

WWN zoning creates zone sets using the WWNs of the attached nodes. This allows you to move nodes from one switch port to another without having to change the zone configuration. The zone set becomes independent of individual switch ports.




Port Zoning

Host

WWPN = 10:00:00:00:C9:20:DC:40


FC Switch

FC Switch Storage

Fabric

WWPN = 50:06:04:82:E8:91:2B:9E


Port Zone 1 = 21,1; 25,3

Port zoning is based on physical topology and uses the domain ID and port number of the switch. This method allows you to change the HBA without changing any zoning information. However, it does not give you the flexibility to physically move attached nodes between switch ports without having to redefine the port number in the zone set.




Zoning Components

Member #1

Member #2

Member #3

Member #4

Zone #1

Zone #2

Active Zone Set

Old Zone Set #1

Old Zone Set #2

Member #5 Zone #3

Members are attached adapters which can be included in a zone. They include HBAs and Storage Fibre adapter ports. A Storage Fibre adapter is a front end port on a storage system. Examples include the Symmetrix FA, CLARiiON SP, and Compaq adapter ports. A zone contains a set of members that can access each other. A member can be a Storage Fibre or Host Bus adapter that is logged into a switch device. Devices spread throughout multiple switches in a multi-switch fabric may be grouped into the same zone. Multiple zones can be assigned membership in zone set. Zones are built and then included into a specific zone set based on the customer’s requirement to activate multiple zones simultaneously. Zone sets are then activated to simultaneously activate the zones in the set.

Zoning a host’s HBA with a Symmetrix or CLARiiON Storage Fibre adapter port allows communication between those members. Volumes visible on that storage port are now available to the host. Members of a zone can see each other; members in different zones cannot.

A zone set is a group of zones that you can activate or deactivate as a single entity in a single or multi-unit fabric. Only one zone set can be active at one time per fabric.

The active zone set is a single zone set that is currently enabled. When a zone set is active, all zones that are members of that zone set are active. Only one zone set can be active for the fabric at one time.




VSANVSANs are logical SANs providing isolation among physically connected devices

VSANs are logical SANs over a common physical fabric allowing logical segments in the network. Multiple independent SANs over a common physical infrastructure provide isolation among devices that are physically connected.




Distance Extension OptionsWDM– Metro region (100s km)– Shared infrastructure for many applications– High bandwidth needs

SONET– Long-distance solutions (1000s km)– Smaller sites feeding into WDM

IP– Native IP for simplicity– Bridged IP when WDM and SONET not available– Reuse part of existing infrastructure

WDMMetro region (100s km)Shared infrastructure for many applicationsHigh bandwidth needsLimitation: Availability of dark fiber; Can be high cost

SONETLong-distance solutions (1000s km)Smaller sites feeding into WDMLimitation: Cost varies (IP can be less expensive); Can be difficult to provision

IPNative IP for simplicityBridged IP when WDM and SONET not availableReuse part of existing infrastructureLimitation: Need a well-designed network; Limited interoperability between IP bridges




SAN Routing - SolutionSAN Routing

A study group dedicated for the development of the standards

Benefits:– Hierarchical network– Local Fabrics can be kept small – Only the shared resources can be shared (routed) between Fabrics

SAN Routing is a new technology development of the FC protocol. A study group within T11 is dedicated to the standards development to accommodate the solution.




SAN Virtualization - SolutionSAN Virtualization

Uses the storage network to redirect IOs on the fly

Benefits:– Free space can be utilized from any/all arrays in the network– Data migration becomes a real time activity

SAN Virtualization uses the storage network to redirect IOs on-the-fly, thereby allowing free space that can be used from any array in the network and makes data migration a real-time activity.




iSCSINative TCP/IP protocol

An IP-based protocol for establishing and managing connections between IP-based storage devices, hosts, and clients

No Fibre Channel content

Today, there are three block storage over IP approaches: iSCSI, FCIP, and iFCP.

iSCSI is a native IP-based protocol for establishing and managing Connections between IP-based storage devices, hosts, and clients. There is no Fibre Channel content.




FCIPTCP/IP based tunneling/encapsulating protocol for connecting/extending Fibre Channel SANs

More IP content, little Fibre Channel content

FCIP is a TCP/IP based tunneling/encapsulating protocol for connecting/extending Fibre Channel SANs using more IP content and less Fibre Channel content.




iFCPGateway to gateway protocol for FC over IPMapping natively in IP across Fibre Channel and IPAn IP-based tunneling protocol for interconnecting Fibre Channel devices together in place of Fibre Channel switchesMore Fibre Channel content; when iFCP creates the IP packets, it inserts information that is readable by network devices and routable within the IP network. IFCP wraps Fibre Channel data in IP packets but maps IP addresses to individual Fibre Channel devices

iFCP is a gateway-to-gateway protocol for FC over IP and maps natively in IP across Fibre Channel and IP. It is an IP-based tunneling protocol for interconnecting Fibre Channel devices together in place of Fibre Channel switches.

When iFCP creates IP packets, it inserts information that is readable by network devices and routable within the IP network. iFCP wraps Fibre Channel data in IP packets, but maps IP addresses to individual Fibre Channel devices.




Block Storage over IP Solutions - NativeAll Ethernet (No Fibre Channel)iSCSI ProtocolEthernet Switches & Routers

LAN LAN

Native iSCSI allows for all communications using Ethernet. Initiators may be directly attached to iSCSI Targets or may be connected using standard Ethernet routers and switches.




Block Storage over IP Solutions - BridgingServers Ethernet AttachedStorage FC Attached (SAN or DAS)iSCSI Protocol

LAN SAN

iSCSI SAN Switch

iSCSI Storage Port

Bridging architectures allow for the Initiators to exist in an Ethernet environment while the storage remains in a Fibre Channel SAN.




Block Storage over IP Solutions - ExtensionServers & Storage SAN AttachedFCIP or iFCP ProtocolSRDF

SAN SAN

WAN

FCIP Routers or iFCP Switches

Extension architectures are most often used to provide connectivity across large distances. Either FCIP or iFCP bring the long distance benefits of IP to Fibre Channel.




Course SummaryKey points covered in this course:

Fibre Channel Architecture

Fibre Channel Layers

Fibre Channel Topologies

SAN concepts

These are the key points covered in this training. Please take a moment to review them.

This concludes the training.

Introduction to Fibre Channel

Documents