Dynamic information storage or retrieval – Condition indicating – monitoring – or testing – Including radiation storage or retrieval
Reexamination Certificate
1999-06-15
2001-08-07
Edun, Muhammad (Department: 2651)
Dynamic information storage or retrieval
Condition indicating, monitoring, or testing
Including radiation storage or retrieval
C369S053150, C369S047140
Reexamination Certificate
active
06272085
ABSTRACT:
TECHNICAL FIELD
This invention relates to a data recording and/or reproducing method and a data recording and/or reproducing apparatus adapted for recording and reproducing digital data with respect to plural disc units.
BACKGROUND ART
In recent years, with realization of multi-channel structure of information offer resulting from popularization of CATV (Cable Television), etc., there is being increased such a demand to record, at the same time, plural video/speech (sound) data, reproduce such data or reproduce such data while recording them from a single video/speech data recording and/or reproducing unit, etc. Further, in order to satisfy this demand, there is being popularized apparatuses called video server (or also called AV (Audio and/or Video) server adapted for recording and/or reproducing video (image) or speech (sound) by using random accessible recording medium such as hard disc, etc.
In general, the video server within the broadcasting station is required to have large capacity because transfer rate of required data is high and long time data is recorded from the requirement with respect to the picture quality and/or the sound quality. In view of the above, there have been carried out an attempt in which data recording and/or reproducing unit including plural hard disc (hereinafter referred to as HD) units adapted for storing video/speech data and capable of carrying out parallel processing is used to thereby realize high speed of transfer rate of data and large capacity, and an attempt in which parity data are recorded to thereby have ability of ensuring reliability even if any HD unit is caused to be out of order. Thus, there can be realized multi-channel video server which can cope with a wide variety of use forms such that even in the case where there are different numbers of channels required by the content of the program or the broadcasting form that the broadcasting station attempts to provide, material data consisting of plural speech/video data are recorded in a distributed manner to carry out, at the same time, sending of many channels, and/or the same material data is reproduced by m any channels with the reproduction time being shifted to thereby construct system such as VOD (Video On Demand) or NVOD (Near Video On Demand), etc.
In the data recording and/or reproducing apparatus used in such video server, there is used RAID (Redundant Arrays of Inexpensive Disks) technology constituted by plural hard disc drives (hereinafter each referred to as HDD) each consisting of plural HDs lectured by the article (“A Case for Redundant Arrays of Inexpensive Disks (RAID)”, ACM SIGMOND Conference, Chicago, Ill., Jun. 1-3, 1988) announced by Patterson, etc. 1988.
In the above-mentioned article, RAID) is classified into five disc systems of RAID-1 to RAID-5. The RAID-1 is the system of writing the same content intc two HDDs. The RAID-3 is the system in which input data is divided into data of a predetermined length to record these data into plural HDDs, and to generate parity data which are exclusive logical sum of data blocks corresponding to each other of respective HDDs to write them into other respective one HDDs. Further, the RAID-5 is the system in which respective divisional units (blocks) of data are caused to be large to record respective divisional data into respective HDDs as respective data blocks, and to record results obtained by taking exclusive logical sum of data blocks corresponding to each other of respective HDDs (parity data) into other respective HDDs as parity blocks to distribute those parity blocks into other respective HDDs.
FIG. 1
is a block diagram showing the fundamental configuration of RAID. This RAID
10
is of configuration comprising a SCSI protocol controller (hereinafter referred to as SPC)
1
, a striping section
2
, a unification section
3
, a memory section
4
, a SPC section
5
, a HDD section
6
, and a CPU
7
. In this example, the HDD section
6
is composed of eight HDDs of HDD
1
~HDD
8
. By taking, as an example, the case where SCSI (Small Computer System Interface) is employed as input/output interface of data, the configurations and the functions of respective sections will be described below.
The SPC
1
is input/output control means for carrying out management/control of input/output of data of the RAID
10
, and its operation is controlled by the CPU
7
. In addition, the SPCI is connected to an initiator
8
through the SCSI.
The striping section
2
is data distribution means for recording recording data inputted through the SPC
1
in a manner distributed to the HDD
1
~HDD
8
of the HDD section
6
. Parity data are respectively added to data distributed to the respective HDDs.
The unification section
3
is data constructing means for constituting data which have been read out from the HDD
1
~HDD
8
of the HDD section
6
into single reproduction data.
The memory section
4
is buffer memory means for temporarily storing recording and/or reproduction data of the HDD section
6
, and is constituted by using semiconductor memory elements.
The SPC section
5
is transfer control means for controlling data transfer between the HDD section
6
and the memory section
4
, and is controlled by the CPU
7
similarly to the SPC
1
.
The HDD section
6
is composed of eight HDDs of HDD
1
to HDD
8
as described above. For these HDDs, small diameter discs having diameter of 3.5 inches or diameter of 5.25 inches are ordinarily used. The same addresses are assigned (allocated) to the respective HDDs as described later. Data are recorded and/or reproduced in parallel.
The memory section
4
and the SPC section
5
both comprise memory
1
~memory
8
and SPC
1
~SPC
8
corresponding to the HDD
1
~HDD
8
of the HDD section
6
. Further, data input/output operations of respective HDDs of the HDD section
6
are carried out through the memory section
4
and the SPC section
5
.
Moreover, the initiator
8
is connected to the SPC
1
through the SCSI, and serves to issue respective various control commands to control the operation of the RAID
10
.
At the time of recording input data into the RAID
10
, parity codes are added to data inputted through the SPC
1
at the time of striping at the striping section
2
. The parity code added data are distributed to the respective HDDs of the HDD section
6
through the SPC section
5
. Moreover, at the time of reproducing data from the RAID
10
, data are read out from the respective HDDs of the HDD section
6
by the procedure opposite to that at the time of recording, and reproduction data are constructed at the unification section
3
.
In HDs of the HDD
2
~HDD
8
data are recorded in the state equally distributed in units called sector. Namely, in the RAID
10
of
FIG. 1
, data are equally distributed to respective sectors of HDD
1
~HDD
8
of the HDD section
6
. Moreover, since the HDD section
6
takes the configuration of the above-described RAID-3 or RAID-5, even if any fault takes place in any one of the HDD
1
~HDD
8
of the HDD section
6
so that normal data transfer cannot be carried out, necessary data are restored by using parity data from data recorded in a manner distributed to the remaining seven HDDs, thus making it possible to normally transfer data. Thus, also in the case where any fault takes place in one HDD, the RAID
10
can continue use without stopping the function.
However, in the case where there take place any faults in the same address of two HDDs or more among these HDD
1
~HDD
8
, there are instances where it would be impossible to restore necessary data to normally transfer data. For this reason, when a fault occurs in one HDD, there is carried out a defect processing to restore the data which could not be read out by fault before fault takes place in the second HDD to record, for a second time, hereinafter referred to as re-record as occasion may demand restored data into the normal alternative sector. This defect processing is carried out by “Reassign Block command” from the CPU
7
.
FIG. 2
is a view for explaining the state where the defect processing i
Edun Muhammad
Frommer William S.
Frommer Lawrence & Haug LLP.
Savit Glenn F.
Sony Corporation
LandOfFree
Data recording/reproducing method and device does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with Data recording/reproducing method and device, we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Data recording/reproducing method and device will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-2442419