GB2168176A

GB2168176A - Addressing data stored on multiple discs

Info

Publication number: GB2168176A
Application number: GB08529623A
Authority: GB
Inventors: Fredericus Henricus Feldbrugge
Original assignee: Philips Gloeilampenfabrieken NV
Current assignee: Koninklijke Philips NV
Priority date: 1984-12-05
Filing date: 1985-12-02
Publication date: 1986-06-11
Also published as: NL8403690A; JPS61134906A; GB8529623D0; DE3542628A1; FR2574211A1

Abstract

A method of addressing data in such a manner that the seek time is reduced in a storage device which includes at least two data carriers, corresponding data being stored in corresponding positions on the storage surfaces of the data carriers (shadow discs), and each data carrier being provided with a separate read member. According to this method, for each data carrier there is determined the actual difference value between the instantaneous position of the associated read member and a newly presented address at which data is to be read. Subsequently, it is determined which difference value has the smallest absolute value and the read member is activated for which the difference value is smallest. Thus, the selected read member need be displaced only over a comparatively small distance so that the seek time is reduced. <IMAGE>

Description

SPECIFICATION Method of addressing data stored in a storage device and storage device for performing the method The invention relates to a method of addressing data stored in a storage device which includes at least two individually drivable data carriers, the corresponding data being stored in corresponding positions on the storage surfaces of the data carriers which storage device also includes an individual read member for each data carrier, a first address being presented to the storage device after which one of the read members and its associated data carrier are displaced relative to one another so as to be positioned at said first address. A method of this kind is known from (an English translation of) the Japanese Patent Application No. 55-135955.The known method is used in a storage device which utilizes a data carrier in the form of two storage discs which can independently rotate about their respective spindles. The two storage discs contain identical data which is stored at corresponding track addresses. Each storage disc has its own read member. When data stored at a first address of the storage device is addressed, both read members are activated and displaced to the first tack address presented. As soon as the read members arrive at the first address presented, they each despatch an acknowledge signal to the control unit. The read member whose acknowledge signal is received first, is instructed to execute the read operation. The seek time for a given address in the storage disc is thus reduced.

It is a drawback of the known method that the recurrent displacement of the two read members consumes extra energy and that, when the addressing operation covers only a limited number of addresses, both read members will be situated in substantially the same position for a next read operation. Moreover, the recurrent displacement of the two read members represents an additional load for the drive mechanism of each read member.

The invention has for its object to realize a method in which the seek time for a given address of the memory is reduced without simultaneous displacement of all read members being required.

To achieve this, a method in accordance with the invention is characterized in that prior to an addressing operation the instantaneous position of each read member, given by second addresses, relative to its associated data carrier is compared with the first address in order to form a series of actual difference values from which the smallest difference value is selected, said positioning operation being performed for the read member data carrier combination for which the difference value is smallest.Because an actual difference value is determined for each read member and the associate data carrier, which difference value represents the distance over which the relevant read member and/or its data carrier must actually be displaced, and because sub sequentiy the smallest actual difference value is determined from all actual difference values, only one read member and/or data carrier will be activated, that is to say that one which has to be displaced over the smallest distance, while the others are not displaced. The determination of these difference values and the selection of the smallest difference value requires hardly any time (a few micro-instructions). The selected read member and/or data carrier is thus displaced over a short distance only, so that the seek time is reduced.

It is to be noted that the determination of actual difference values between a first address presented and a series of second addresses is known per se from the Japanese Patent Application No. 57-166659. However, the storage device described therein comprises only one data carrier and the data stored therein can be read by a plurality of read members. Said actual difference value is calculated for each read member and the read member which must actually travel the shortest distance in order to reach the first address receives the read instruction.The method known from the Japanese Patent Application No. 57-166659 differs from a method in accordance with the invention in that the known method utilizes only one data carrier, while a method in accordance with the invention utilizes at least two data carriers on which corresponding data is stored, each data carrier having its own read member. The use of a plurality of read members for the same data carrier has the drawback that in the case of a head crash between one of the read members and the data carrier, the data carrier is liable to be damaged to such an extent that it becomes almost impossible to read reliable data therefrom by means of the other read members.

In a preferred version of a method in accordance with the invention the data at the first address is read after completion of said positioning operation. This preferred version is characterized in that on the basis of a predetermined error recognition criterion it is tested whether the data read is reliable, a read member data carrier combination other than the combination for which the difference value is smallest being selected when the data is deemed to be unreliable. The advantages of the reduced seek time and the use of a plural- ity of data carriers (shadow discs) are thus combined.

When the storage device includes at least three data carriers, said other selected read member/data carrier combination is preferably the combination with which the smallest difference value but one is associated. The seek time is thus minimized also when another read member/data carrier combination is selected.

The invention also relates to a storage device which includes at least two individually drivable data carriers, the corresponding data being stored in corresponding positions on the storage surfaces of the data carriers, which storage device also includes an individual read member for each data carrier, positioning means for displacing each read member/data carrier combination relative to one another so as to be positioned at a first address presented, and comparison means. Such a storage device is also known from the Japanese Patent Application No. 55-135955.

A storage device in accordance with the invention is characterized in that the comparison means are suitable to compare, under the control of a first address received, the first address with an instantaneous position, given each time by a second address, of each read member relative to its associated data carrier, and to form actual difference values between the first address and a respective second address, which storage device includes selection means which is suitable to select the smallest actual difference value from a series of actual difference values received and to generate a selection signal for the read member/data carrier combination with which the smallest difference value is associated, said positioning means being suitable to control, under the control of a selection signal received, the positioning operation for the read member/data carrier combination indicated by the selection signal received. Such a storage device enables a method in accordance with the invention to be carried out.

A preferred embodiment of a storage device in accordance with the invention is characterized in that the data carriers are formed by optical or magnetic discs. The seek time in optical or magnetic storage discs is comparatively long, so that a seek time reduced in accordance with the invention leads to a substantially reduced access time for reading such a storage disc, and hence a more efficient use of the storage device.

The invention will be described in detail hereinafter with reference to the drawing.

Therein: Figure 1 shows an embodiment of a storage device in accordance with the invention; Figure 2 shows an embodiment of a control unit which forms part of a storage device in accordance with the invention, and Figure 3 shows a flowchart illustrating the method in accordance with the invention.

The invention will be described with reference to an embodiment in which the data carrier is formed by two storage discs. However, it will be apparent that the invention can be used in any storage device which comprises data carriers in which the corresponding data is stored at corresponding positions in the storage surfaces of the data carriers. Data carriers of this kind are, for example disc-shaped or drum-shaped memories as well as slide-like (magneto-optical) memories.

Fig. 1 shows an embodiment of a storage device in accordance with the invention. The storage device (G) includes a reading device and a data carrier which includes two storage discs 5 and 6 in the present embodiment. The storage discs are formed, for example by two optical or magnetic discs on which the data is stored in tracks. The storage discs 5 and 6 contain identical data in corresponding memory locations so-called shadow discs or mirrored discs. Shadow discs are known per se and are used inter alia for error tolerances.

Because identical data is stored on two different discs, the risk of incorrect data being supplied by the storage device is substantially reduced. The storage discs 5 and 6 are arranged on spindles 9 and 10, respectively.

Each of the spindles 9 and 10 forms part of a drive device (not shown in the Figure). Each disc is thus driven by its known drive device.

This is important in view of the fact the data on the two discs must be independently accessible and that a disturbance on one storage disc may not influence the other storage disc.

The data stored in the storage discs 5 and 6 is read by a read member 7, 8, respectively, which is connected to the reading device 2 via a connection line 3, 4, respectively.

The reading device 2 is connected to a central unit 1. The central unit 1 fcirms part of, for example a data processing system whose storage device is a peripheral station. The central unit inter alia despatches read instructions to the storage device in order to fetch data stored in the storage discs for further processing.

Theread device 2 includes a control unit 40 for controlling the read members 7 and 8. Fig.

2 shows a preferred embodiment of such a control unit. The control unit includes a first register 20 and a second register 21 for storing the instantaneous position of the read members 7 and 8, respectively. A third register 22 is provided for the storage of a first track address which is presented via the input 34 and which forms part of a read instruction issued by the central unit. This is because such a read instruction usually contains a track address and a sector address which together form a memory address at which data is to be read. The third register 22 includes three data outputs; a first and a second data output are connected to a data input of the first (20) and the second (21) register, respectively. A thid data output of the register 22 is connected to a first arithmetic input of a subtraction unit 23, 24 respectively. A data output of the first (20) and the second (21) register is connected to a second arithmetic input of the substraction unit 23, 24, respectively. Each of the subtraction units 23, 24 determines the difference value between the numerical values presented to their first and second arithmetic inputs. The arithmetic output of each of the subtraction units is connected to a respective absolute value unit (25, 26) which determines the actual difference value of the difference value determined by the subtraction units. An output of each of the absolute value units 25 and 26 is connected to a respective input of a comparison unit 27 which determines the smallest difference value from the difference values presented to its two inputs.A result output of the comparison unit 27 is directly connected to the connection line 3 and to a first gate input of a logic OR-gate 30 and, via an inverter 28, to the connection line 4 and a first gate input of a logic OR-gate 31. The control unit is controlled by a unit 29 which applies a first control pulse to the line 32 which is connected to a second gate input of the logic OR-gates 30 and 31 and to a control input of the third register 22. A second control pulse is applied to line 33 which is connected to a control input of the two absolute value units 25 and 26.

When a new first track address is presented to the input 34, it is stored in the third register 22. The control unit 29 which is also connected to the input 34 generates the first control pulse after having detected the presentation of a first rrack address. Via the logic ORgates 30 and 31, the first control pulse is applied to the first (20) and the second (21) register. Under the control of this first control pulse, the instantaneous positions of the read members 7 and 8, given by a respective second track address, are applied to the second arithmetic inputs of the subtraction units 23 and 24, respectively. Again under the control of the first control pulse, the first track address stored in the third register is applied to the first arithmetic inputs of the substraction units 23.The subtraction units determine the difference value between the first and the second track address presented. Under the control of the second control signal, the difference values are transferred to the absolute value units 25 and 26 in order to determine the absolute value thereof. When the read members are displaced in only one direction for the execution of a read operation, this fact is taken into account by the absolute value units, for example by adding a fixed value which covers a full stroke of the read member to a negative difference value. These actual difference values are subsequently compared in the comparison unit 27.In the present embodiment the comparison unit 27 determines whether the difference value (A,) on the output if the absolute value unit 25 is smaller than the difference value (2) on the output of the absolute value unit 26. If At tod,, a result signal having the logic value "1" is output on the result output of the comparison unit. Via the connection line 3, this logic "1" is applied to the read member 7 which is thus activated.

This is because when A, tod,, the read member 7 whose instantaneous position was stored in the register 20 is situated nearest to the new track address so that it may execute the read operation because the read member 7 will require the smallest amount of time in order to reach said second track address.

Via the logic OR-gate 30 the result signal having the logic value "1" is also applied to the first register 20 which stores, under the control thereof, the new track address from the third register 22 at which the read member 7 will be situated after execution of the read operation. Because the inverter 28 translates the value of the result signal into a logic "0", neither the read member 8 nor the second register 21, will be activated.

However, if A, > A2, the read member 8 is situated nearest to the new track address and will execute the read operation. The comparison unit then generates a result signal having the logic value "0". This logic value "0" is converted into a logic "1" by the inverter 28, and activates the read member 8 as well as the second register 21 in which a new second track address is now loaded.

When both read members are situated in a corresponding position, At=52. In that case the comparison unit generates a result signal having the logic value "1", but it will be apparent that this is merely a choice and that the result signal could just as well have the logic value "0".

The determination as to which of the read members 7 or 8 will have to travel the shortest distance in order to reach the new track address presented can also be performed under the control of the central unit 1. The central unit then executes a processing program.

Fig. 3 shows a flowchart illustrating such a processing program. The processing program is started (50) after the central unit has issued a read instruction which inter alia contains a first track address. The first track address (SA) is stored in a first internal register of the central unit (51). Subsequently, the instantaneous position of the read member 7, given by a second track address P7, is taken up (52).

This is realized, for example, by reading from a memory location associated with the central unit or the read member 7 in which said second track address P7 is stored. Subsequently (53) the central unit determines the actual difference value between the second track address P7 and the first track address SA received (SA-P7=At). This difference value (tut) is stored in a second internal register of the central unit (54). Subsequently, the instantaneous position of the read member 8, given by a second track address P8, is taken up (55) (again by reading from a memory location).

The central unit then determines (56) the actual difference value SA-P8=A2 and stores (57) this difference value 2 in a third internal register. Subsequently (58), the central unit determines which of the difference values A1 and 2 is smaller. This is because the smallest difference value represents the shortest distance to be travelled by one of the read members. If A1#A2, the read member 7 is activated (59) by means of an activation signal which is generated by the central unit and which is applied to the read member 7 via the connection line 3. The new first track address reached by the read member 7 after termination of the read operation is stored in the relevant memory location (60).However, if A1 > A2, the read member 8 will be activated (61) and the new first track address will be stored in the relevant memory location in an analogous manner. After storage of the new track address, the processing program is terminated (63).

The read member which is situated nearest to the newly presented track address is thus displaced to the new track address, so that the seek time is reduced.

It will be apparent that the execution of a method in accordance with the invention is not restricted to the described versions. In a further version of a method in accordance with the invention, the storage device may include more than two storage discs which all contain identical data in corresponding memory locations. When there are N (No2) of such storage discs which all have respective read members, the difference A, (1 cisN) is determined for each of these N read members and the read member for which A, has the smallest value is activated.The smallest difference value A, is determined, for example by first determining all values A, and by determining the smallest value A, from the set thus formed, or by comparing each nearly formed difference value A, with a preceding difference value (A,#) which has a smaller absolute value and by substituting A, for A" when 4 < A".

Let us assume that on the basis of the above selection criterion (A1,A2) the storage disc 5 and the associated read member 7 (Fig. 1) are selected. The data present at the track address SA is then read. This data is subsequently tested for reliability on the basis of the error recognition criterion (for example, CRC). When the data is reliable, it will be further processed. However, if the data is not reliable because it does not satisfy the recognition criterion, it will be rejected and subsequently the read member 8 will be positioned at the address SA. The read operation is then performed in the storage disc 6 by the read member 8.

When N > 2 and unreliable data is read from the selected storage disc, the smallest value but one of A, is determined from the series of difference values A,. The read member/storage disc combination associated with the smallest value but one of A, is then selected to execute the read operation. The selection of the smallest value but one of 4 is performed, for example by storing the various calculated values of A1 in a storage element according to increasing magnitude and by addressing the storage locations in succession each time when a next value of A1 must be fetched for the same read operation at the address SA.

Claims

1. A method of addressing data stored in a storage device which includes at least two individually drivable data carriers, the corresponding data being stored in corresponding positions on the storage surfaces of the data carriers, which storage device also includes an individual read member for each data carrier, a first address being presented to the storage device after which one of the read members and its associated data carrier are displaced relative to one another so as to be positioned at said first address, characterized in that prior to an addressing operation the instantaneous position of each read member, given by second addresses, relative to its associated data carrier is compared with the first address in order to form a series of actual difference values from which the smallest difference value is selected, said positioning operation being performed for the read member/data carrier combination for which the difference value is smallest.

2. A method as claimed in Claim 1 in which, after said positioning operation, the data is read at said first address, characterized in that on the basis of a predetermined error recognition criterion it is checked whether the data read is reliable, a read member/data carrier combination other than the combination for which the difference value is smallest being selected when the data is deemed to be unreliable.

3. A method as claimed in Claim 2, in which the storage device includes at least three data carriers, characterized in that said further selected read member/data carrier combination is the combination with which the smallest difference value but one is associated.

4. A storage device which includes at least two individually drivable data carriers, the corresponding data being stored in corresponding positions on the storage surfaces- of the data carriers, which storage device also includes an individual read member for each data carrier, positioning means for displacing each read member/data carrier combination relative to one another so as to be positioned at a first address presented, and comparison means, characterized in that the comparison means are suitable to compare, under the control of a first address received, the first address with an instantaneous position, given each time by a second address, of each read member relative to its associated data carrier, and to form actual difference values between the first address and a respective second address, which storage device includes selection means which is suitable to select the smallest actual difference value from a series of actual difference values received and to generate a selection signal for the read member/data carrier combination with which the smallest difference value is associated, said positioning means being suitable to control, under the control of a selection signal received, the positioning operation for the read member/data carrier combination indicated by the selection signal received.

S. A storage device as claimed in Claim 2, characterized in that data carriers are formed by optical or magnetic discs.

6. Comparison means for use in a storage device as claimed in Claim 4, characterized in that the comparison means includes an address input for receiving the first address and a register memory for storing the second addresses.

7. A method of addressing data stored in a a storage device, substantially as hereinbefore described with reference to the accompanying drawings.

8. A storage device for performing the method claimed in Claim 7, substantially as hereinbefore described with reference to the accompanying drawings.