Electrical computers and digital processing systems: memory – Storage accessing and control – Specific memory composition
Reexamination Certificate
1997-06-06
2001-01-23
Chan, Eddie P. (Department: 2751)
Electrical computers and digital processing systems: memory
Storage accessing and control
Specific memory composition
Reexamination Certificate
active
06178480
ABSTRACT:
A patent application entitled METHOD OF OPERATING A DISK STORAGE SYSTEM has been filed for Fouad A. Tobagi, Joseph M. Gang, Jr., Randall B. Baird, Joseph W. M. Pang, and Martin J. McFadden on Nov. 17, 1992, bears Ser. No. 07/977,493, now U.S. Pat. No. 5,581,784 and is assigned to the assignee hereof. Another patent application entitled VIDEO APPLICATION SERVER has been filed for James E. Long, Joseph M. Gang, Jr., Charles J. Bedard, Randall B. Baird, and David A. Edwards on Jun. 24, 1993, bears Ser. No. 08/082,227, now U.S. Pat. No. 5,550,932 and is assigned to the assignee hereof. The above-identified applications contain subject matter related to the subject matter of the present application and are incorporated herein by reference.
FIELD OF THE INVENTION
The present invention relates to a method for operating a disk storage system which stores video data so as to maintain the continuity of a plurality of video streams. In particular, the present invention provides a method for operating a disk storage system which is flexible with respect to storage capacity, streaming capacity, the start up latency of new streams, and the amount of buffer capacity required. The inventive method also has advantageous properties with respect to scalability (the addition of more disks), reliability, and multiple bit rates.
BACKGROUND OF THE INVENTION
The demand for networked digital audiovisual systems is expected to grow considerably over the next few years as businesses, government and other institutions increasingly turn to digital networks to distribute audiovisual information for education, presentations and reference applications. These customers expect systems that will allow a number of people to be able to view audiovisual information from a server simultaneously, while fully retaining their other network functions.
The characteristics of files, file access and network traffic in digital video applications differ substantially from those encountered in data applications. With data applications, whenever a user makes a file access request to a server, or requests that data be transmitted on a network, the user expects a fast response—fast compared to the time it takes it to place the next request. As a result, the capacity of a server and the overall bandwidth must both be large compared to the average demand placed by a single user. Accordingly, the design of a file server aimed at supporting data applications and the design of a network to support data traffic have been based on the principle of bandwidth sharing and statistical time multiplexing. File servers have furthermore taken advantage of the property of locality in file access, and incorporated appropriate caching mechanisms. In all cases, as the overall load on the shared resources increased, the average response time experienced by all users also increased.
Consider now digital video. A video signal is analog in nature and continuous over time. It is digitized by first sampling it at regular intervals, and then by quantizing each sample. This digitization process results in a data stream which is of relatively constant and very high rate (NTSC signals result in data rates in the neighborhood of 100 Mb/s and an HDTV signal, 600 Mb/s.) However, given that the sampled data exhibits a great deal of redundancy, compression is applied, thus significantly reducing the stream's rate. Depending on the bandwidth of the original analog signal, the sampling rate, the quantization step size, the encoding method, and the desired image quality, the resulting data rate for a digital video signal ranges from 64 Kb/s to tens of Mb/s. For example, CCITT Recommendation H.261 specifies video coding and decoding methods for audio visual services at the rate of px64 Kb/s, where p is in the range of 1 to 30 (i.e., 64 Kb/s to 2 Mb/s). The MPEG standard specifies a coded representation that can be used for compressing video sequences to bit rates around 1.5 Mb/s. Advances in compression techniques and their VLSI implementations are among the important reasons why video services over LANs and WANs are becoming practical.
Two important observations can be made. The first is that the volume of bits corresponding to a digitized video segment of useful duration (even compressed) is large. A ten minute MPEG video segment requires over 100 Mbytes of storage; ten hours requires over 5 Gbytes. Thus video servers where such video information is to be stored must have relatively large storage capacity.
The second observation is that the communication of digital video data between two nodes on a local area network (e.g., a server and a desktop station) requires that data be transmitted in a stream fashion. This means that video data packets must be delivered to the destination on time, and failure to deliver video data packets on time would result in video quality degradation. This has two main implications: (i) from a network's point of view, one requires the availability, on a continuous basis, of a bandwidth at least equal to the signal's data rate; (ii) from a file and storage system point of view, one requires streaming capabilities which guarantee the continuity of each stream being retrieved or stored. Thus, in order to support multiple independent video signals, the network must have the necessary aggregate bandwidth as well as means to guarantee the bandwidth required for each video stream, and the file storage system must be of the streaming type and must have a capacity sufficient to handle all video streams. By the same token, there is a maximum number of video steams of a given data rate that a network and a server can support, and means are provided to prevent additional requests from overloading the system.
It is thus clear that the characteristics of video traffic differ substantially from those of traditional data traffic to the point that servers and local area networks designed primarily to support data applications are not appropriate to effectively support video services. New capabilities in servers and networks must be offered.
The focus in this application is on the design and operation of a storage system suitable for a video server, noting again that the storage requirements for video data are different from the storage requirements for typical LAN data in two respects:
(i) the size of video files is an order of magnitude greater or more; even with compression techniques, the physical storage needs are large.
(ii) when serving a video stream, be it for recording or playback, it is desirable to maintain the continuity of the stream. In the case of playback, data must be retrieved from the storage medium, transmitted over the network, and made available to the decoder no later than the time at which it is needed so as to avoid letting the decoder run dry. Similarly, when a stream is getting recorded, the writing of data on the storage medium must keep up with the rate at which it is getting produced so as to avoid buffer overflow and data loss.
A server's storage system preferably satisfies the following requirements:
(i) provide random access capability for both recording and playback;
(ii) have the storage capacity required by the application;
(iii) have the I/O throughput required to support simultaneously a target maximum number of users (streams); and
(iv) guarantee a start-up latency for new streams within a specified maximum tolerable threshold.
Due to their random access read and write capability, wide availability, and low cost, magnetic disk drives emerge as very appropriate storage components for this purpose. Multiple drives can be used in order to provide the necessary storage capacity and/or to achieve an aggregate throughput larger than that achieved with a single drive.
A magnetic disk storage system
20
is illustrated in FIG.
1
. The disk storage system
20
comprises a plurality of disk drives
200
. Each disk drive
200
comprises a disk
21
and a controller
210
. The disk drive
200
is shown in greater detail in FIG.
2
. As shown in
FIG. 2
, the disk
21
of the disk dive
200
comprises a
Baird Randall B.
Gang, Jr. Joseph Mark
Pang Joseph W. M.
Tobagi Fouad A.
Chan Eddie P.
Ellis Kevin L.
Proskauer Rose LLP
Starlight Networks, Inc.
LandOfFree
Method for operating an array of video storage units does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with Method for operating an array of video storage units, we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Method for operating an array of video storage units will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-2516739