Electrical computers and digital data processing systems: input/ – Input/output data processing – Input/output data buffering
Reexamination Certificate
1998-11-17
2001-08-21
Lee, Thomas (Department: 2182)
Electrical computers and digital data processing systems: input/
Input/output data processing
Input/output data buffering
C710S007000
Reexamination Certificate
active
06279057
ABSTRACT:
This application is related to the following applications filed on even data herewith, each of which is incorporated by reference: “Method and Dedicated Frame Buffer for Loop Initialization and Responses” by Judy Lynn Westby, “Method and Apparatus for Using CRC for Data Integrity in On-Chip Memory” by Judy Lynn Westby, and “Method and Apparatus to Reduce Arbitrated-Loop Overhead” by Judy Lynn Westby and Michael H. Miller.
FIELD OF THE INVENTION
The present invention relates to the field of mass-storage devices. More particularly, this invention relates to an improved fibre-channel arbitrated-loop (“FC-AL”) apparatus and method having dedicated frame buffers for receiving frames.
BACKGROUND OF THE INVENTION
One key component of any computer system is a device to store data. Computer systems have many different devices where data can be stored. One common place for storing massive amounts of data in a computer system is on a disc drive. The most basic parts of a disc drive are a disc that is rotated, an actuator that moves a transducer to various locations over the disc, and circuitry that is used to write and read data to and from the disc. The disc drive also includes circuitry for encoding data so that it can be successfully retrieved from and written to the disc surface. A microprocessor controls most of the operations of the disc drive, in addition to passing the data back to the requesting computer and taking data from a requesting computer for storing to the disc.
The interface for transferring data between the disc drive and the rest of the computer system is typically a bus or channel, such as the Small Computer Systems Interface (“SCSI”), or the Fibre Channel. Certain aspects of such interfaces are often standardized in order that various devices from different manufacturers can be interchanged and all can be connected to a common interface. Such standards are typically specified by some standards committee of an organization such as the American National Standards Institute (“ANSI”).
One standardized interface for exchanging data between various storage devices and various computers is the fibre channel. In some embodiments, the fibre-channel standard includes arbitrated loops (described further below). In some embodiments, the fibre-channel standard supports a SCSI-like protocol for controlling data transfers.
Fibre channels represent significant advantages over Small Computer Standard Interface (“SCSI”) designs. Fibre channels provide significantly higher bandwidths, currently up to about 106 megabytes per second, compared to between two and twenty megabytes per second for traditional SCSI designs. Fibre channels provide greater connectivity in that up to one-hundred twenty-six devices (including the host) may be connected, as compared to a maximum of seven or fifteen devices in typical SCSI environments. The fibre channel can be attached with a single connector and does not require a switch. A fibre channel using coaxial electrical conductors operates at distances of up to thirty meters between devices, and up to ten kilometers using fibre optics for an entire channel, as compared to a maximum total length of up to twenty-five meters for SCSI environments. In SCSI environments, errors in data transmission are detected through use of parity, whereas in fibre channels, errors are identified by a running disparity and cyclic-redundancy-code check (“CRC check”) information. More information can be found in U.S. Pat. No. 5,802,080 entitled “CRC Checking Using a CRC Generator in a Multi-port Design,” and U.S. Pat. No. 5,663,724 entitled “16B/20B Encoder,” both by the present inventor, Westby, and commonly assigned to the present assignee Seagate Technology, Inc., each of which is incorporated by reference.
The fibre-channel arbitrated loop (“FC-AL”) is an industry-standard system employing a byte-oriented DC-balanced (0,4) run-length-limited 8B/10B-partitioned block-transmission code scheme. The FC-AL operates at a clock frequency of 106.25 MHZ. One form of an 8B/10B encoder/decoder is described in U.S. Pat. No. 4,486,739 granted Dec. 4, 1984 for “Byte Oriented DC Balanced (0,4) 8B/10B Partitioned Block Transmission Code” by Franaszek et al., which is incorporated by reference.
A fibre-channel arbitrated loop (“FC-AL”) allows for multiple devices, each called “a node,” to be connected together. A node may be any device (a computer, workstation, printer, disc drive, scanner, etc.) of the computer system having an interface allowing it to be connected to a fibre-channel “topology” (defined just below). Each node has at least one port, called an NL port (“node-loop port”) to provide access to other nodes. The components that connect two or more ports together are collectively called a “topology” or a “loop.” Each node communicates with all other nodes within the provided topology or loop.
Ports are the connections in a fibre-channel node, though which data may pass over the fibre channel to ports of other nodes (the outside world). A typical fibre-channel drive has two ports packaged within the drive's node. Each port includes a pair of “fibers”—one to carry information into the port and one to carry information out of the port. Each “fiber” is a serial data connection, and, in one embodiment, each fiber is actually a coaxial wire (e.g., coaxial copper conductors, used when the nodes are in close proximity to one another); in other embodiments, a fiber is implemented as an optical fiber for at least some of its path (e.g., when nodes are separated by an appreciable distance, such as nodes in different cabinets or, especially, different buildings). The pair of fibers connected to each port (one carrying data into the port, the other carrying data out from the port) is called a “link” and is part of each topology. Links carry information or signals packaged in “frames” between nodes. Each link can handle multiple types of frames (e.g., initialization, data, and control frames).
Since each fiber carries data in one direction only, nodes are connected to one another along a loop, wherein the nodes must arbitrate for control of the loop when they have data to transfer. “Arbitration” is the process of coordinating the nodes to determine which one has control of the loop. Fibre-channel arbitrated loops attach multiple nodes in the loop without hubs or switches. The node ports use arbitration operations to establish a point-to-point data-transfer circuit. FC-AL is a distributed topology where each port includes at least the minimum necessary function to establish the circuit. The arbitrated-loop topology is used to connect any number of nodes between two and one-hundred twenty-six (126) node ports.
In some embodiments, each node includes dual ports (each connected to a separate loop) which provide redundancy, so that if one loop fails, the other one can fulfill the loop duties. Dual ports also allow two hosts (e.g., two host computers) to share a single drive.
In typical first- and second-generation FC-AL drives, the two ports shared the frame-validation and frame-generation logic. This meant that if one port was receiving or transmitting a frame, the alternate port was effectively busy (since it could not simultaneously use the frame-validation and frame-generation logic), and the alternate port was thus forced to deny its host-bus adapter permission to send frames. Some host-bus adapters would continuously have to arbitrate and attempt to send a frame over and over until the primary port closed. Also, the drive was only able to transmit on one port at a time. In some cases, an outbound data transfer on a given port would have to be paused in order to send a response or perform loop initialization on the other (alternate) port.
CRC Background
Most data-transmission operations employ error checking by which an error code, based on the header and payload data of the transmission, is checked to verify the integrity of the received header and payload data. One such error-checking scheme employs cyclic-redundancy-code (“CRC”) information. A typical circuit employing CRC error che
Du Thuan
Lee Thomas
Schwegman Lundberg Woessner & Kluth P.A.
Seagate Technology Inc.
LandOfFree
Communications system having dedicated frame buffers located... does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with Communications system having dedicated frame buffers located..., we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Communications system having dedicated frame buffers located... will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-2446819