CS 635 Lecture 6, Part 1

Fibre Channel

Designed to combine the speed of I/O channel communication with the flexibility and interconnectivity of protocol-based network connection.

I/O Channel oriented facilities

Data type qualifiers for routing frame payload into particular interface buffers
Link-level constructs associated with individual I/O operations
Protocol interface spec to allow support of existing I/O channel architecture, such as SCSI

Network oriented facilities

Full multiplexing of traffic between multiple destinations
Peer-to-peer connectivity between any pair of ports on a fibre channel network
Capabilities for internetworking to other connection techniques

Fibre Channel is 1 Gbps transfer interface technology that maps several common transport protocols including IP and SCSI allowing it to merge high-speed I/O and internetworking functionality in a single connectivity technology.
Fibre channel is a open standard as defined by ANSI and OSI standards and operates over copper and fiber optic cabling at distances up to 10 km.

It supports multiple interoperable topologies including point-to-point, arbitrated loop and switching and it offers several qualities of service for network.

Fibre Channel Architecture

Full duplex with two fibers per link
Performance from 100 to 800 Mbps on a single link (200 to 1600 Mbps per link)
Support for distances up to 10 km
Small connectors

Fibre Channel Elements

The network consists of nodes and one or more switching elements. The collection of switching elements is called a fabric. These elements are interconnected by point-to-point links between ports on the individual nodes and switches. Communication consists of the transmission of frames across the point-to-point links.

XXX picture 9.2

Each node includes three or more ports, called N_Ports, for interconnection. Each fabric switching element includes one or more ports, called F_Ports. Any node can communicate with any other node connected to the same fabric using the services of the fabric. All routing of frames between N_Ports is done by the fabric. Data may be buffered within the fabric, making it possible for different nodes to connect to the fabric at different data rates. Fabric is responsible for buffering and routing frames between source and destination nodes.

Fibre channel fabric is not a shared medium, therefore there are no MAC related issues. Fibre Channel scales easily in terms of N_Ports, data rates, and distance covered.

Fibre Channel Protocol Architecture

XXXX picture 9.3
The Fibre Channel standard is organized into five levels.

FC-0 Physical media

Optical cable with laser or LED transmitters
Copper coaxial cable for highest speed over short distances
Shielded twisted pair for lower speeds over short distances

Currently, data rates ranging from 100 to 800 Mbps per fiber are defined. Depending on the data rate and medium involved, max distances for individual point-to-point links range from 50 meters to 10 km.

FC-1 Byte synchronization and encoding

8B/10B encoding/decoding scheme provides balance, is simple to implement, and provides useful error-detection capability. 8 bits of data from level FC-2 is converted into 10 bits for transmission. The 8B/10B scheme was patented by IBM.
Special code character maintains byte and word alignment

FC-2 Actual transport mechanism

Framing protocol and flow control between N_Ports
Tree classes of service between ports

This level deals with the transmission of data between N_Ports in the form of frames. Among the concepts defined at this level:

Node and N_Port and their identifiers
Topologies
Classes of service provided by the fabric
Segmentation of data into frames and reassembly
Grouping of frames into logical entities called sequences and exchanges
Sequencing, flow control, and error control.

FC-3 Common service layer

Port related services
Services across two or more ports in a node

FC-3 provides a set of services that are common across multiple N_Ports of a node. The functions include

Striping: makes use of multiple N_Ports in parallel to transmit a single information unit across multiple links simultaneously. This achieves higher aggregated throughput. For transferring large data sets in real time, as in video-imaging applications
Hunt groups: a set of associated N_Ports at a single node. This set is assigned an alias identifier that allows any frame set to this alias to be routed to any available N_Port within the set. This may decrease latency by decreasing the chance of waiting for busy N_Port.
Multicast: delivers a transmission to multiple destinations. This includes sending to all N_Ports on a fabric (broadcast) or to a subset of the N_Ports on a fabric.

FC-4 Upper layer protocols

Supports a variety of channel and network protocols

FC-4 defines the mapping of various channel and network protocols to FC-PH. I/O channel interfaces include:

Small computer system interface (SCSI)
High-performance parallel interface (HIPPI)

Network interfaces include:

IEEE 802: frames map onto Fibre Channel frames
Asynchronous Transfer Mode (ATM)
Internet Protocol

Physical Media and Topologies

Transmission Media

Optical Fiber Transmission Media

The standard specifies both single-mode and multimode optical fiber alternatives. Optical fiber will operate up to 1.0625 Gbps and distances up to 10 km. S

Coaxial Cable Transmission Media

Lower cost alternative to optical fiber. Supports up to 1.0625 Gbps up to 24m., and 266 Mbps up to 47m.

Shielded Twisted Pair

Used only over very short distances.

Topologies

Fabric or switched topology is the most common. Includes at least one switch to interconnect a number of N_Ports. (figure 9.5). The fabric topology may include a number of switches forming a switched network.

Routing in the fabric topology is transparent to the nodes. Each node in the configuration has a unique address. When data from a node are transmitted into the fabric, the edge switch to which the node is attached uses the destination port address in the incoming data frame to determine the destination port location. The switch then either delivers the frame to another node attached to the same switch or transfer the frame to an adjacent switch to begin the routing of the frame to a remote destination.

Good scalability:

additional nodes increase aggregate capacity of the network
the fabric is protocol independent and largely distance insensitive
the switch maybe changed without affecting the overall configuration
burden on nodes is minimized: each node is only responsible for managing a point-to-point connection between itself and the fabric

Point-to-point topology includes two N_Port nodes connected to each other with no fabric switch. There is no routing
Arbitrated loop topology is low cost topology for connecting up to 126 nodes in a loop. The ports on the arbitrated loop must contain the functions of both N_Ports and F_Ports; these are called NL_Ports. Roughly equivalent to the token ring protocol. This topology was developed with peripheral connectivity in mind. It natively maps SCSI.

Framing Protocol

The FC-2 framing protocol defines the rules for the exchange of higher-layer information between nodes. FC-2 specifies types of frames, procedures for their exchange, and formats. FC-2 is similar to the data link layer of OSI.

Classes of Service

Fibre Channel provides a logical system of communication called Class of Service. Five different Classes of Services are available:

Class 1 - Acknowledged connection service
Class 2 - Acknowledged connectionless service
Class 3 - Unacknowledged connectionless service
Class 4 - Fractional bandwidth connection oriented service
Class 6 - Simplex connection service

Class 1 Service

Provides dedicated path though the fabric, similar to circuit-switched network. Provides guaranteed data rate between two communicating ports and guaranteed delivery of frames in the order in which they are transmitted.
To initiate class 1 service, a N_Port transmits a frame with a special start-of-frame delimiter SOFc1. This alerts the fabric that a connection is requested. The fabric allocates a circuit between the requesting N_Port and the destination N_Port. The destination N_Port can then transmit an ACK indicating its acceptance of the connection to the requesting N_Port.
Class_1 is used for large files and absolute quality of service.

Class 2 Service

Connectionless service, independently switching each frame and providing guaranteed delivery with an acknowledgment of each receipt. The path between two interconnected devices is not dedicated. The switch multiplexes traffic from N_Ports and NL_Ports without dedicating a path through the switch.

Class 2 credit-based flow control eliminates congestion that is found in many connectionless networks. If no buffer space is available, a "Busy" is sent to the originating N_Port. The N_Port will then re-send the message. This way no data is arbitrarily discarded just because the switch is busy at the time.

Typical Class 2 frame latency is less than one microsecond, making it ideal for shorter data transfers like those in most business applications.

Class 3 Service

Class 3 is a connectionless service, similar to Class 2, but no confirmation of receipt is given. This unacknowledged transfer is used for multicast and broadcasts on networks and for storage interface on Fibre Channel loop. The loop establishes a logical point-to-point conniption and reliably moves data to and from storage.

Class 3 arbitrated loop transfers are also used for IP networks. Some applications use logical point-to-point connections without using a network layer protocol, taking advantage of Fibre Channel's reliable data delivery.

Class 4 Service

Class 4 is fractional bandwidth, connection-oriented service. Virtual connections are established with bandwidth reservation for a predictable quality of service (Qos). A class 4 connection is bi-directional with one virtual circuit operation in each direction and supports a different set of QoS parameters for each VC (guaranteed bandwidth and latency). A node may reserve up to 256 concurrent Class 4 connections.

Class 4 flow control is end-to-end and provides guaranteed delivery. Class 4 is ideal for time-criticial applications like video.

Class 6 Service

Class 6 is similar to class 1, providing simplex connection service. However, class 6 also provides multicast and preemption. Class 6 is ideal for video broadcast application and real-time systems that move large quantities of data.

Intermix

Intermix mode allows the reservation of full Fibre Channel bandwidth for a dedicated (class 1) connection but also allows connectionless traffic within the switch to share the link during idle Class 1 transmission. An ideal application for Intermix is linking multiple large file transfers during system backup. During a Class 1 file transfer, a Class 2 or 3 message can be sent to the server to set up the next transfer. Upon completion of one transfer, the next will immediately start, increasing efficiency.

Top Next