WO2009112964A1

WO2009112964A1 - Method and apparatus for locating information on storage media

Info

Publication number: WO2009112964A1
Application number: PCT/IB2009/050799
Authority: WO
Inventors: Hermann Zach; Peter Maria Franciscus Rombouts
Original assignee: Nxp B.V.
Priority date: 2008-03-12
Filing date: 2009-02-27
Publication date: 2009-09-17

Abstract

A method of locating information on storage media is provided, wherein the method comprises providing a plurality of identification pairs, wherein each pair is formed by an identification label for a specific information block and a corresponding actual storing location label of the specific information block, and estimating an internal state indicative of approximated locations of a plurality of information blocks by fitting the plurality of identification pairs using a fit algorithm.

Description

METHOD AND APPARATUS FOR LOCATING INFORMATION ON STORAGE MEDIA

FIELD OF THE INVENTION

The invention relates a method of locating information blocks on storage media, in particular to a method of efficiently locating information on linear storage media.

Beyond this, the invention relates to an apparatus for locating information on storage media, in particular to an apparatus for efficiently locating information on linear storage media.

Moreover, the invention relates to a computer readable medium. Furthermore, the invention relates to a program element.

BACKGROUND OF THE INVENTION Linear storage media, such as magnetic storage tapes, are devices which data is written to or read from by moving a read/write head relative to the medium in a linear motion, i.e. the head is moved over the medium or conversely the medium is moved over the head.

Repositioning the head from one location on the medium to another can take a relatively long time depending on the physical characteristics of the medium like e.g. the length of the medium, its inertia, limitations on the velocities and forces that can be applied, etc.

In general the data access time on a linear storage medium can be divided in two parts corresponding to the following operations as schematically shown in Fig. 1. A first part is the re-positioning of the read/write head from a start position 100 to an intermediate position 101 at high speed and without reading from the medium, a so-called fast forward operation. A second part comprises the reading from the medium until a marker or identifier is reached indicating the start of the requested data at position 102, a so-called linear read operation.

In the prior art several solutions exist to improve the access times of linear storage media. For example, LTO magnetic tapes use a partial index, i.e. they divide the medium in parts called "wrap sections" and store the record numbers of the data records last written to each "wrap section" on the cartridge memory (CM) module. These stored record numbers are subsequently used to determine in which "wrap section" a record is stored, when it is requested for retrieval. Evidently the more index entries are kept the better the intermediate position will approach the actual position. US 7,111,135 describes a method to increase the resolution of the index kept by LTO tapes and provide redundancy by incorporating two or more cartridge memory (CM) modules. GB 2 393 804 describes a method to optimise the total read time of a set of data blocks by determining an optimal read sequence to read these data blocks, while US 2003/0026021 describes a method to alter the layout of a linear tape medium so as to provide faster access to parts of the stored data.

OBJECT AND SUMMARY OF THE INVENTION

However, there may be a need to provide an alternative method of locating information on storage media, an alternative apparatus for locating information on storage media, a computer readable element, and a program element exhibiting an improved performance. In order to meet the need defined above, a method of locating information on storage media, an apparatus for locating information on storage media, a computer readable medium and a program element according to the independent claims are provided. Further improvements are disclosed in the dependent claims.

According to an exemplary aspect of the invention a method of estimating an internal state indicative of locations of information on storage media is provided, wherein the method comprises providing a plurality of identification pairs, wherein each pair is formed by an identification label for a specific information block and a corresponding actual storing location label of the specific information block, and estimating an internal state indicative of approximated locations of a plurality of information blocks by fitting the plurality of identification pairs using a fit algorithm.

In particular, the fit algorithm may be a non-linear fit algorithm and/or the storage media is a linear storage media, like a magnetic tape. In particular, all or only a subset of the plurality of identification pairs may be used in the fit. Moreover, the internal state may be stored in a storing device, e.g. in the storage medium device or at a device external to the storage medium device. Another possible storing device for the internal state may be an RFID- tag or transponder, which may also be adapted to perform the method according to the above described exemplary aspect of the invention. In case the internal state is stored in a device external to the device performing the above mentioned method, the internal state may be exported to the external device.

According to an exemplary aspect of the invention a method of locating information blocks on storage media is provided, wherein the method comprises providing an identification label, and estimating an approximated location based on an internal state and the identification label. In particular, the internal state may be an internal state which is estimated according to a method of providing an internal state indicative of locations of information on storage media according to an exemplary aspect of the invention.

According to an exemplary aspect of the invention an apparatus for providing an internal state indicative of locations of information on storage media is provided, wherein the apparatus comprises a processing unit, wherein the processing unit is adapted to receive a plurality of identification pairs, wherein each pair is formed by an identification label for a specific information block and a corresponding actual storing location label of the specific information block, and wherein the processing unit is further adapted to estimate an internal state relating to approximated locations of a plurality of information blocks by fitting the plurality of identification pairs using a fit algorithm.

In particular, the apparatus may perform all operations involving the internal state. Moreover, the internal state may be stored internally or may be exported for storage outside of the apparatus. According to an exemplary aspect of the invention an apparatus for locating information blocks on storage media is provided, wherein the apparatus comprises a processing unit, wherein the processing unit is adapted to receive an identification label, and wherein the processing unit is further adapted to estimate an approximated locations based on an internal state and the identification label. According to an exemplary aspect of the invention an information storage system is provided which comprises an apparatus for locating information blocks on storage media according to an exemplary aspect of the invention and a storage device adapted to read/write the information blocks on a storage media, wherein the apparatus and the storage device are adapted to communicate with each other. In particular, the storage device may be a read/write unit adapted to read/write on a linear storage media, e.g. may be a tape drive which is adapted to write/read a magnetic tape. According to an exemplary aspect of the invention a program element is provided, which, when being executed by a processor, is adapted to control or carry out a method according to an exemplary aspect of the invention.

According to an exemplary aspect of the invention a computer-readable medium is provided, in which a computer program is stored which, when being executed by a processor, is adapted to control or carry out a method according to an exemplary aspect of the invention.

In particular, the information block may be formed by a data block and may relate to stored information or to information written to a storage medium, while the location label may be indicative for the position the corresponding information or data block is written on the storage medium.

The term "identification pair" may particularly denote a pair of labels and/or numbers, wherein one of the labels or numbers characterizes a specific information or data block, while the other one characterizes the location of the corresponding specific information or data block stored on a storage medium, e.g. a meter counting or winding counting of a storage tape.

The term "internal state" may particularly denote a compact representation of all pairs of identification label (ID) and corresponding actual location (A), i.e. like a complete index, but allowing a small margin error on the position, so that in general only estimated or approximated positions (E) may be extracted from the internal state but not the actual positions or locations (A). Permitting a margin of error however may allow encoding the internal state much more compactly. Such an "internal state" may be represented by a mathematical function or relation, e.g. the coefficients of a polynomial fit. Furthermore, the internal state may be estimated in such a way that it is possible to derive an estimated storage position of a specific data block or information block based on the unique identifier or identification label of the specific data block and the internal state.

By providing a method estimating or calculating such an internal state, which may be used to approximate respective storage positions of the stored data blocks or information blocks on a specific device or storage media, less memory space may be necessary compared to a full or complete index as known from the prior art. In particular, it may not be necessary, when using a method according to an exemplary aspect of the invention, to keep an index of the identification pairs, so that disadvantages inherently coupled to such an index may be overcome. In particular, it may not be necessary to increase the needed memory space to such an extend as in the prior art, in which, when keeping a complete index for improving the estimated position, the necessary storage space linearly increases with the number of indexed pairs. Thus, memory space may be saved independent of whether the internal state is stored on the storage media, e.g. linear storage media, or on an external memory. Furthermore, the updating operation may be easier when using a method according to an exemplary aspect of the invention compared to a method using a (full) index which will require frequent updates/queries and hence should at least during reading/writing operations be stored on a medium with good random access characteristics, which may be obsolete or at least the necessary random access memory space may be reduced when using a method according to an exemplary aspect of the invention. Some further disadvantages of the prior art which may be overcome by the present invention are:

• In case the index is stored on the linear storage medium: o If it is not cached it will require frequent repositioning of the read/write head to read or update it. o If it is cached the size of the cache should be big enough to accommodate the entire index. o Furthermore if it is cached the cached and stored indexes will have to remain synchronised requiring regular updates or at least resulting in increased load and unload times. o Updating the complete index requires changing data already stored on the linear storage medium - the linear storage medium's file system should support this.

• In case the index is stored outside the linear storage medium: o The size of the external storage medium should be big enough to accommodate the entire index. o The external storage medium should have a file system that supports updating stored items. o The external storage should have sufficient bandwidth. If, in case of prior art methods, the index is kept small to avoid the problems mentioned above this leads to other disadvantages:

• The partial index does not keep information on the data blocks located between the indexed data blocks; i.e. it has low granularity. o Estimates derived from it will be less accurate and can lead to longer average random data access times due to imprecise positioning.

• Updates/queries of the partial index will still be frequent and require good random access characteristics of the medium that stores the index. • Current methods do not allow choosing which index entries to keep such that the entries that are kept can have very low information content.

At least some of the disadvantages of the prior art mentioned above may be overcome by a method and an apparatus according to an exemplary aspect of the invention.

The usage of a method according to an exemplary aspect of the invention may particularly result in a reduced time to locate a data block or information block on a storage media, e.g. a linear storage media, in the reduction of loading and unloading times, in reduced storage requirements, and good granularity that may furthermore be fine-tuned to the application.

A gist of an exemplary aspect of the invention may be to provide a method of locating information on a linear storage medium, wherein the method may be dividable into two sub-methods. In one sub-method a first, parametrizable algorithm is provided that estimates a position or location from an identifier or identifier label based on an internal state, e.g. on the values of parameters of the internal state. That is, this sub-method may calculate an estimated storage position of information based on an identifier of that information and a fitted internal state. This internal state may be estimated by a second sub-method. In the second sub- method a second algorithm is provided which estimates an internal state, e.g. representing an optimal set of parameters, based on identifiers and actual storage positions. The two sub- methods may be implemented as two separate methods or by a method combining both sub- methods. Next, further exemplary embodiments of the method of locating information blocks on storage media are described. However, these embodiments also apply to the apparatus for locating information blocks on storage media, the system for locating information blocks on storage media, the computer-readable medium, and the program element. According to another exemplary embodiment of the method the fit algorithm is a polynomial fit algorithm. In particular, the internal state may be represented by coefficients of the polynomial fit. In particular, the internal state may be represented by all or only a subset of the fitted coefficients. According to another exemplary embodiment of the method the fit algorithm is adapted in such a way that the approximated storage location of a specific information block represents an earlier location than the actual storage location of the specific information block.

In particular, the algorithm may be adapted such that each approximated or estimated storage location (E) represents an earlier location (underestimating) than the actual storage location or position (A). However, it is also possible that the algorithm is only adapted in such a way that such earlier locations are only highly preferred over later locations

(overestimating) but that such later locations are not prohibited or prevented for sure. Such a preferring may be ensured by introducing different weight factors for underestimating and overestimating the actual storage location.

The term "earlier location" may particularly denote a location or position on a linear or serial storage media, like a magnetic tape, which is reached earlier by a reading head when the reading head starts from the beginning of the linear storage media.

In the art many algorithms are known to encode data sets, in this case the pairs of the identification label and the corresponding actual positions, into an internal state. Different algorithms may lead to different internal states and different trade-offs. Some of these trade-offs may include trade-offs between several factors such as: the accuracy of the estimated position, the memory space needed to store the internal state, the time required to derive the estimated position from the internal state, in particular on a specific device, the time required to update the internal state, in particular on a specific device.

In general, the fit algorithm may be adapted in such a way that the resulting internal state which can be used to calculate an estimated position (E) ensures that the estimated position is always before the actual position (A) of the respective data block or information block, so that the estimated position may serve as a starting point for a linear reading operation (E < A), and that the estimated position approximates the actual position (E - A). According to another exemplary embodiment the method further comprises providing a further identification pair, and updating the internal state based on an additional fit taking into account the further identification pair. In particular, the fit for the updating may be performed using the same or another fit algorithm. Furthermore, the updating may be based on one or more further identification pairs.

Next, further exemplary embodiments of the apparatus for locating information blocks on storage media are described. However, these embodiments also apply to the method of locating information blocks on storage media, the system for locating information blocks on storage media, the computer-readable medium, and the program element.

According to another exemplary embodiment the apparatus further comprises a device for reading/writing the storage media. In particular, the apparatus which is adapted to perform the estimation process may be the storage device itself, e.g. may be implemented into a magnetic tape writing/reading device or even may be joined to a magnetic tape cartridge, e.g. an RFID transponder.

According to another exemplary embodiment the apparatus further comprises a communication interface adapted to communicate with a storage device. That is, the apparatus may be external to the storing device, e.g. to the storing device handling the writing/reading of the storage media. For example, the apparatus may be implemented in an RFID transponder, like the one present in LTO magnetic tapes. The apparatus may store the state internally and may be joined to the linear storage medium, e.g. integrated into a packaging or cartridge. Alternatively, the apparatus may be integrated in a reader/writer device, for example, and may not store the state internally but exports it to a separate device, e.g. the linear storage medium, a further storage device attached to the linear storage medium, or a server. Furthermore, the apparatus may be formed by a set of two or more sub-devices, one sub-device joined to the linear storage medium that is only used to derive estimated positions of a data block based on a stored internal state received from the second external sub-device that calculates or estimates the internal state.

By providing an apparatus adapted for the method which is external to the storage device it may be possible to increase the flexibility of the implementation, to improve a forward/backward compatibility in case of system changes, and interoperability. For that it may be possible to implement a standardised interface, e.g. comprising a standardised interface terminal and a standardised protocol, which can be provided to the operations that work on the internal state, e.g. to operations that update and derive an estimated position from the internal state. Such a standardised interface may allow abstracting the implementation including the internal state. Summarizing, a gist of an exemplary aspect of the invention may be seen to provide a method of locating information on storage media which estimates a storage position of a data block on a linear storage media wherein the estimation is based on a fit algorithm using identification pairs (ID/A) as fitting parameters, wherein the fit algorithm may be adapted in such a way that the estimated storage position or location is lower or earlier than the actual storage location. Such a method may provide for a more close approximation of the actual location than current techniques known in the art and may be more convenient to use. In particular, it should be noted that the closer the estimated storage position, which is used as a start position for the linear reading, is to the actual storage location the higher the performance of random data retrieval may be.

The apparatus may use any existing function known in the art or a function specifically developed for this purpose to determine an estimated position from the internal state and/or to determine the internal state from a data set containing the identification pairs (ID/A). If, for example, the identifier (ID) is a numeric value a polynomial approximation can be used to calculate the estimated position form the identifier. In that case the internal state may consist of the coefficients of the polynomial and can be calculated by a constrained polynomial curve fitting.

A method or an apparatus according to an exemplary aspect may be applied to improve the access times of all linear storage media. It may primarily valuable for enhancement to current LTO tape cartridges and drives.

The aspects and embodiments defined above and further aspects of the invention are apparent from the examples of embodiment to be described hereinafter and are explained with reference to these examples of embodiment. It should be noted that features described in connection with a specific exemplary embodiment or exemplary aspect may be combined with another exemplary embodiment or another exemplary aspect.

BRIEF DESCRIPTION OF THE DRAWINGS

The invention will be described in more detail hereinafter with reference to examples of embodiment but to which the invention is not limited.

Fig. 1 schematically illustrates access time components of a linear storage media. Fig. 2 schematically illustrates the operation of an apparatus according to an exemplary embodiment.

Fig. 3 shows an example of an approximating function.

DESCRIPTION OF EMBODIMENTS

The illustration in the drawing is schematic. In different drawings, similar or identical elements are provided with similar or identical reference signs.

Fig. 2 schematically illustrates the operation of an apparatus according to an exemplary embodiment. Fig. 2 schematically depicts a linear storage medium 201 having a data track 202 thereon on which an estimated position (E) and an actual position (A) are schematically shown as lines 203 and 204, respectively. Fig. 2 further shows an apparatus or device (D) 205 adapted to perform a method according to an exemplary embodiment of the invention. The device 205 may be an RFID transponder and may have stored an internal state

(S) 206 on it and/or may be adapted to calculate the internal state (S) according to a method of the present invention and is adapted to output the estimated position (E) based on the a respective inputted identifier (ID) of the requested information. In general, the RFID transponder may implement, e.g. by software, hardware or a hybrid way, a method of providing an internal state indicative of locations of information blocks on storage media according to an exemplary aspect of the invention and/or the RFID transponder may implement a method of locating information blocks on storage media according to an exemplary aspect of the invention. For example, the RFID transponder may hardcode an algorithm to calculate an estimated storage position of specific information from a received internal state, e.g. set of parameters, based on a received identifier of the specific data. The

RFID transponder may further hardcode an algorithm to estimate the internal state itself based on identification pairs of identifiers (ID) and respective actual storage positions (A).

As mentioned the device may be an external storage device, e.g. is an RFID transponder integrated into a linear storage mediums cartridge. The linear storage medium is read from and written to by a drive which communicates to the device trough an RF channel.

The RFID transponder therefore implements at least two commands used by the drive:

1. SEND: Used by the drive to send information on the data block identifiers (ID) and the corresponding actual positions (A) to the transponder (D). 2. GET: Used by the drive to get an estimated position (E) for a data block identified by a data block identifier (ID) from the transponder (D).

During writing operations the drive uses the SEND command to send the data block identifiers (ID) and the corresponding actual positions (A) on the linear storage medium to the transponder (D) that uses it to update its internal state (S).

During read operations when a particular data block needs to be located the drive will use the GET command to send the corresponding data block identifier (ID) to the transponder (D). The transponder will use this identifier and the internal state (S) to calculate an estimated position (E) on the tape where the (linear) read operation will start. Alternatively to the storing of the internal state in an RFID transponder, the storing may be implemented on the linear storage medium drive in another embodiment. In this embodiment the device (D) is a linear storage medium drive that can store the internal state (S) to e.g. the linear storage medium or to an external memory like an RFID transponder attached to the storage medium. During writing operations the drive updates the internal state (S) using data block identifiers (ID) and the corresponding actual positions (A) on the linear storage medium, or alternatively the drive calculates the internal state (S) based on a full index of data block identifiers (ID) and corresponding actual positions (A) after the writing operation has completed. When no further updates of the internal state are needed; e.g. when the medium is unloaded from the device it is stored, either on the linear storage medium or on an external storage device.

During read operations when a particular data block needs to be located the drive will read back the internal state (S) and use it to together with the corresponding identifier (ID) of the data block to calculate an estimated position (E) on the tape where the (linear) read operation will start.

Fig. 3 schematically illustrates an example of an approximation function. In particular, Fig. 3a depicts a graph 301 representing identification pairs of data or information blocks as position in bytes versus the data index or identification label, i.e. Fig. 3a shows the position in bytes versus data index. Fig. 3b depicts a graph 302 representing an approximated function corresponding to an internal state as well as the graph 301 shown in Fig. 3a. The approximated function reflects the graph 301 of identification pairs of data blocks versus the data index shown in Fig. 3 a quite well. If for example the identifier (ID) is a numeric value a polynomial approximation can be used to calculate the estimated position (E) form the identifier (ID). In that case the internal state (S) will consist of the coefficients of the polynomial and can be calculated by a constrained polynomial curve fitting. The approximation function can furthermore be designed such that it takes all prior knowledge of physical layout (e.g. tracks) and the ordering of the data on the linear storage medium into account. This will further reduce the storage requirements of the internal state (S).

Finally, it should be noted that the above-mentioned embodiments illustrate rather than limit the invention, and that those skilled in the art will be capable of designing many alternative embodiments without departing from the scope of the invention as defined by the appended claims. In the claims, any reference signs placed in parentheses shall not be construed as limiting the claims. The word "comprising" and "comprises", and the like, does not exclude the presence of elements or steps other than those listed in any claim or the specification as a whole. The singular reference of an element does not exclude the plural reference of such elements and vice-versa. In a device claim enumerating several means, several of these means may be embodied by one and the same item of software or hardware. The mere fact that certain measures are recited in mutually different dependent claims does not indicate that a combination of these measures cannot be used to advantage.

Claims

1. A method of providing an internal state indicative of locations of information blocks on storage media, the method comprising: providing a plurality of identification pairs (ID, A), wherein each pair is formed by an identification label (ID) for a specific information block and a corresponding actual storing location label (A) of the specific information block, estimating an internal state (S) indicative of approximated locations (E) of a plurality of information blocks by fitting the plurality of identification pairs (ID ,A) using a fit algorithm.

2. The method according to claim 1, wherein the fit algorithm is a polynomial fit algorithm.

3. The method according to claim 2, wherein the internal state (S) is represented by coefficients of the polynomial fit.

4. The method according to claim 1, wherein the algorithm is adapted in such a way that the approximated storage location (E) of a specific information block represents an earlier location than the actual storage location (A) of the specific information block or highly prefers an earlier location over a later location.

5. The method according to claim 1, further comprising: providing a further identification pair, and updating the internal state based on an additional fit taking into account the further identification pair.

6. A method of locating information blocks on storage media, the method comprising: providing an identification label (ID), and estimating an approximated location (E) based on an internal state (S) and the identification label.

7. An apparatus for providing an internal state indicative of locations of information blocks on storage media, the apparatus comprising: a processing unit, wherein the processing unit is adapted to receive a plurality of identification pairs (ID, A), wherein each pair is formed by an identification label (ID) for a specific information block and a corresponding actual storing location label (A) of the specific information block, and wherein the processing unit is further adapted to estimate an internal state (S) relating to approximated locations (E) of a plurality of information blocks by fitting the plurality of identification pairs (ID, A) using a fit algorithm.

8. The apparatus according to claim 7, further comprising: a device for read/write the storage media.

9. The apparatus according to claim 7, further comprising: a communication interface adapted to communicate with a storage device.

10. An apparatus for locating information blocks on storage media, the apparatus comprising: a processing unit, wherein the processing unit is adapted to receive an identification label (ID), and wherein the processing unit is further adapted to estimate an approximated locations (E) based on an internal state (S) and the identification label (ID).

11. An information storage system, comprising: an apparatus for providing an internal state indicative of locations of information blocks on storage media according to claim 7, and a storage device adapted to read/write the information blocks from/to a storage media, wherein the apparatus and the storage device are adapted to communicate with each other.

12. The information storage system according to claim 11, further comprising: an apparatus for locating information blocks on storage media according to claim 10.

13. A program element, which, when being executed by a processor, is adapted to control or carry out a method according to claim 1 and/or claim 6.

14. A computer-readable medium, in which a computer program is stored which, when being executed by a processor, is adapted to control or carry out a method according to claim 1 and/or claim 6.