JP3430999B2 - Storage medium scheduler, real-time system, and scheduling method - Google Patents

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a technique for scheduling a request for data including stream data to a storage medium, and a technique for scheduling a processing order according to parameters added for the request to the storage medium, and more particularly to When handling stream data and non-stream data in a storage medium, information for scheduling is added to the request to the storage medium, and the request order is set based on the added information so that stream data will not be lost. The present invention relates to a storage medium scheduler, a real-time system, and a scheduling method for normally processing stream data that is time-limited and severely managed by executing request scheduling, and also normally processes non-stream data. .

[0002]

2. Description of the Related Art In a conventional storage medium, the data recorded in the storage medium is almost always time-unconstrained because of the processing capacity and storage capacity of the system and the request processing speed of the storage medium. The normal processing method for request processing is that the requests are processed in the order in which the requests are issued.

Conventionally, since there is no time restriction on the data handled by the storage medium, even if the request processing time is long, if the request is processed normally, no system problem occurs. Therefore, there is no problem with the processing method that processes the requests in the order in which the requests are issued.

However, with the recent increase in capacity and speed of storage media and the improvement in system processing capacity for controlling storage media, it has become possible to handle time-constrained data, which has been difficult to handle until now. The data that requires this time constraint is mainly moving image and audio data. Data that has such temporal continuity is called stream data, but since such stream data is image / sound data, it must be provided stably to provide the data to be reproduced or recorded. Stream data will not be played back / recorded properly because of lack of data.

Next, a case will be described as an example where stream data is handled by the above-mentioned conventional request processing system and a 5-minute stream is composed of five 1-minute stream data. If the start time of the first stream data is 00:00:00 (h: m: s), the second stream data cannot be played normally unless it is read by 00:00. Similarly, the third one is 0
The processing start time of the stream data to be provided starts from the start time of the stream (0:02:00, the fourth one is 00:03:00, the fifth one is 00:04:00, and so on). Recording on the storage medium or reproduction from the storage medium must be completed by the N-th -1) x "stream data time per one". In this example, since all the requested requests are stream data, there is no problem even in the conventional processing method if the storage medium has a band capable of processing the stream data, but a problem occurs if the band is insufficient.

Next, a case where stream data requests and non-stream data requests are mixed in the requests will be described. When 5 requests are issued, the 1st, 2nd, 4th and 5th requests are stream data requests and the 3rd is non-stream data requests, the 1st, 2nd, 4th and 5th requests are stored media as described above. If the bandwidth is sufficient, there will be no problem in playing / recording the stream.

Further, in the case of the conventional processing method, after the requests for the first and second stream data are processed, 3
The second non-stream data request is processed, but if this processing end time has exceeded the fourth stream data playback time, the time lag between the second stream data and the fourth stream data is seen for the entire stream. Will occur.

In addition, in the case of recording, the time lag is lost, and in the case of reproduction, the stream is stopped for the time lag, and a normal stream cannot be constructed.

In the case of stream reproduction, since the stream data is read from the storage medium, the data recorded in the storage medium will not be disturbed. Therefore, if the same stream reproduction is performed again, normal reproduction can be performed. Although it is possible, in the same situation, the normal stream cannot be reproduced again, so that the system is not in a preferable state.

In the case of stream recording, the time lag is fatal. If the stream data recorded on the storage medium is lost, and if the stream data being recorded is data that is provided only once, the missing portion cannot be recorded again, so the stream data being recorded Will be incomplete and will have the worst result for the system.

As described above, the conventional request method for a storage medium is not very suitable for handling stream data having a time constraint. Therefore, in order to handle stream data, it is necessary to perform request scheduling in consideration of real-time property.

[0012]

However, the biggest problem in the method of processing requests in the order of request issuance to the conventional storage medium is the stream data with time constraint and the non-stream with no time constraint. When the request configuration is such that data is mixed, the completion degree of the stream is affected by the request processing time of the non-stream data.

The number of combinations of requests to the storage medium is actually limited, but there are enormous combinations, and a flexible request scheduling method is required in order to normally process non-stream data while normally handling stream data. is necessary. Therefore, it is very difficult to handle stream data and non-stream data in a storage medium in a mixed manner without using a unique request scheduling method so that stream data can be processed normally.

The present invention has been made in view of the above problems, and an object thereof is to provide information for scheduling in a request to a storage medium when stream data and non-stream data are handled in the storage medium. Is added, and request scheduling is performed based on the added information so that stream data is not lost, so that stream data that is time-constrained and difficult to manage can be processed normally. In addition, a storage medium scheduler, a real-time system, and a scheduling method for normally processing non-stream data are also provided.

[0015]

The gist of the invention described in claim 1 is a storage medium scheduler for optimally processing a request for a storage medium, which includes priority, time limit, and
A parameter that identifies the access destination and expiration date inside the storage medium.
For the request with a meter, the storage medium
First queue with priority given to optimization of internal access destination
And a second optimized in the time limit order.
Means for creating a queue of
The priority, the time limit, and the storage as optimized
Based on the parameters of the access destination inside the medium,
Means for planning the execution order for the
Scan the queue and the second queue to find the optimal
Executing the request, the first queue and the second queue
For updating the queue of the request and the execution result of the request
Is transmitted to the request source,
Access the request to the body inside the storage medium.
Queue that prioritizes optimization of the destination and the time limit order
Create two types of queues optimized by
With reference to the two types of queues,
Scan queues that are prioritized for access optimization
However, if there is something that exceeds the time limit, execute the request
In the order of the queue, the order is optimized according to the time limit order
The request has been processed, and the time limit has been increased.
In this case, optimization of the access destination inside the storage medium has priority.
Requests are processed in the order of queue
Has a configuration that switches the est execution order in real time
It consists in a storage medium scheduler, characterized in that. The gist of the invention according to claim 2 resides in the storage medium scheduler according to claim 1 , wherein the storage medium includes video data for multimedia use including video-on-demand. Further, the gist of the invention described in claim 3 resides in the storage medium scheduler according to claim 1 or 2 , wherein the storage medium includes audio data for multimedia use including video on demand. . The storage medium according to claim 1 or 2 , wherein the storage medium includes data other than video and audio for multimedia use including video-on-demand. It resides in the media scheduler. Further, the gist of the invention according to claim 5, claim 1, characterized in that it comprises means for indicating the priority depending on the type of data recorded in the storage medium or
In the storage medium scheduler described in 2 . The gist of the invention according to claim 6 resides in the storage medium scheduler according to claim 1 or 2 , wherein the storage medium is a magnetic storage disk. The gist of the invention according to claim 7 resides in the storage medium scheduler according to claim 1 or 2 , wherein the storage medium is an optical storage disk. The gist of the invention according to claim 8 resides in the storage medium scheduler according to claim 1 or 2 , wherein the storage medium is a magnetic tape. Further, the gist of the invention according to claim 9 is a storage medium.
A schedule for optimally scheduling body requests.
A real-time system having a scheduling means
Thus, the scheduling means waits for optimization of the access destination inside the storage medium prioritized with respect to the request having parameters for specifying priority, time limit, access destination inside storage medium, and expiration date. Matrix and before
The time limit was exceeded in the request that constituted the optimized queue in the order of the time limit and was normally processed in the order of the priority order of the optimization of the access destination in the storage medium. When the request is about to be processed, the processing order is optimized according to the time limit order.
Process the request by changing to a queue
If all of the requests of both the queue does not exceed the time limit, and wherein the scheduling so that the storage medium inside the access destination of the optimization processes in the order of prioritized queues again It exists in a real-time system. Further, the gist of the invention according to claim 10 is that, when the new request is generated, the scheduling means waits prioritizing optimization of an access destination inside the storage medium based on the parameter of the request. Matrix, and optimized by the time limit order
Relative queue resides in real-time system of claim 9, characterized in that it comprises a means for reconstructing the content. The gist of the invention according to claim 11 is 2
11. The real-time system according to claim 10 , further comprising: one queue, and a means for switching a processing order in real time according to a state of the request to be processed. The twelfth aspect of the invention is a scheduling method for scheduling a request execution to the storage medium having parameters for specifying a priority, a time limit, an access destination inside the storage medium, and an expiration date. Is added to the request
Access destination inside the storage medium based on the parameters
To configure the first queue prioritized optimization of
In addition, a second queue optimized in the time limit order is set up.
Process and when a new request is requested,
For the first queue and the second queue
The process of reconstructing and the access destination inside the storage medium
The time limit present in the queue prioritized for optimization
The process to check when processing a request, and the most
Request the time limit present in the optimized queue
And the storage medium to be executed
Optimized access destinations inside the body are placed in a priority queue
If the existing request exceeds the time limit
In addition, the execution order is placed in the optimized queue according to the time limit order.
Optimal according to the process of transferring and the time limit order to be executed
If the request existing in the encrypted queue is the time
Access within the storage medium if it is within the limit
The process of shifting the execution order to the queue that prioritized the previous optimization
And the request you are trying to execute has the expiration date
If it exceeds the limit, immediately stop processing the request
And a scheduling step .

[0016]

BEST MODE FOR CARRYING OUT THE INVENTION The gist of the present invention will be described first.
Generally, in a recording system that handles stream data, “priority”, “time limit”, and “expiration date” are used as request request parameters. In the present invention, it is defined as follows.

Time limit: The time at which processing should be started. Recording / playback does not have to be interrupted even if the time is not in time .

Expiration date: Time at which a part of recording / playback is interrupted unless the processing is started by this time. A request whose expiration date has passed is meaningless for the stream even if it is processed, and is used as a material for determining whether or not the request request can be executed.

Priority: This is the priority of processing the request, but in the present embodiment, it is used as auxiliary information for the determination processing when the previous two parameters are the same.

The present invention is roughly divided into two components. One component is a block characterized by investigating a parameter added to a request request to a storage medium and generating two types of queues, that is, a time limit order and an access destination order inside the storage medium. .

This also has a function of checking the two queues already created and reconstructing each queue again when a new request is made.

The other component investigates the parameter of the request to be processed in the currently used queue among the two created queues, and if the time limit is exceeded, This block is characterized in that the first request is executed from the queue in the time limit order, and if the time limit is not exceeded, the first request is executed from the access destination order queue inside the storage medium.

It is also one of the features of the present invention that the processed request has a function of deleting from the above two queues in order to prevent re-execution of the processed request.

The greatest feature of the present invention is that the above-mentioned two types of queues are alternately used according to the result of examination of the retrieved request.

Next, the operation of the present invention will be described. The request for the storage medium is issued with the information for request scheduling added. The issued request examines the parameters for scheduling, and first creates a queue according to the time limit order and a queue according to the access destination order inside the storage medium. At this time, if the parameters used at the time of creation are the same in either queue,
The priority included in the parameter determines the in-queue order.

When the generation of the two queues is completed, the queue generation block notifies the request execution block of the completion of queue generation. This queue generation completion notification is notified when the request in the queue is registered for the first time, and is not notified when there is one or more requests in the queue.

When a new request is generated, the above-mentioned operation is performed to constantly update the contents of each queue. When the request execution block receives the notification of completion of queue creation, it examines the current usage queue.

The parameters of the request are examined from the queue of the examination result, and when the request exceeds the time limit, the requests are executed in order from the queue in the order of time limitation, and the executed request is the above-mentioned two waits. Delete registered executed requests from the queue.

If this request does not exceed the time limit, the requests are sequentially executed from the access destination in-order queue inside the storage medium, and the executed request is deleted from each queue.

If the request to be executed exceeds the expiration date, the request is meaningless even if it is executed, so the request is not executed because the bandwidth load on the storage medium is reduced and an abnormal termination is requested. Notify the original.

As described above, by alternately using the two queues of the request queue not subject to the time constraint and the request queue in consideration of the time constraint, the time constraint of stream data etc. It is possible to process requests for certain data in the optimal order. Hereinafter, embodiments of the present invention will be described in detail with reference to the drawings.

FIG. 1 is a block diagram of a storage medium scheduler for explaining the first embodiment of the present invention. In FIG. 1, 100 is a queue generation block, 101 is a time limit order queue, 102 is an access destination order queue inside a storage medium, 103 is a request execution block, 104 is a storage medium controller, and 105 is a storage medium. There is.

Referring to FIG. 1, according to the embodiment of the present invention, a storage medium 105, a queue generation block 100, a time limit order queue 101, an access destination order queue 102 inside the storage medium, and an executed request are provided. A request execution block 103 having the function of deleting from the queue,
The storage medium controller 104 is provided.

The queue generation block 100 first sets a time limit based on parameters such as priority, time limit, access destination inside the storage medium 105, and expiration date which are added to the request when the request is issued. Investigate,
The requests are registered in the time limit order queue 101 so that they are processed in the time limit order. Next, the access destination inside the storage medium 105 is investigated, and the request is registered in the access destination order queue 102 inside the storage medium so that the access destinations inside the storage medium 105 are processed in order. At this time, if there is no request in the queue and the request is the first registration request for the current state of the queue, the request execution block 103 is in the state of waiting for request registration, so request notification of request registration. Perform to execution block 103. If there are already registered requests in each queue, all the requests in the queue are reconfigured according to the respective queue conditions by adding the newly registered request.

FIG. 2 shows an example of request order generation in the queue, and shows the inside of the time-limited order queue 101 used by the queue generation block 100 to reconstruct the request in the queue. In FIG. 2, 11
Reference numeral 0 indicates the inside of the queue that was previously generated, 111 indicates the registered request, and 112 indicates the inside of the queue after the sequence regeneration. H added to the request in Figure 2
The numerical value in the form of h: mm: ss represents the time limit of the request.

The previously created queue interior 110 shown in FIG. 2 represents the state before a new request request is issued. At this time, a new request (Fig. 2
When a registered request 111) is issued, the queue generation block 100 reexamines the requests in the queue and regenerates an appropriate order.
In this case, the new request time limit is Request
Since it falls in the middle of st2 and Request3, the regenerated queue becomes the queue internal 112 after order regeneration as shown in FIG. In this case, the time limit order queue 101 is used as an example, but the same applies to the access destination order queue 102 inside the storage medium.

The request execution block 103 examines the queue currently in use after executing the request. As a result of the examination, if there is no request in the queue, the execution is suspended and the state of waiting for registration of a new request is entered.
If there is a request, the first request in the queue is fetched and the process proceeds to request execution processing. The fetched request has been fetched from the same queue as the previously processed request.

The request execution block 103 examines the above-mentioned parameters of the fetched request. If the time limit is exceeded, the request execution block 103 fetches the next request to be executed from the time limit order queue 101, and also exceeds the time limit. If there is no such request, the next request to be executed is fetched from the access destination queue 102 inside the storage medium, and the storage medium controller 104 is requested to access the storage medium 105 to make a request to the storage medium 105.

When the instruction to the storage medium controller 104 is executed, the request execution block 103 deletes the processed request from the above two queues.

Next, referring to the flowcharts of FIGS. 3 and 4, the operation of the storage medium scheduler (scheduling method)
Will be described. 3 and 4 show the queue generation block 1
FIG. 3 is a flowchart showing an outline of the operation of 00 in FIG.
Is the operation of the first half of the processing steps of the queue generation block 100 of FIG. 1, and FIG. 4 is the queue generation block 10 of FIG.
The operation of the latter half of the processing step of 0 is shown.

Referring to FIG. 3, in the present embodiment, first, when a request is registered, the time limit value of the parameters added to the request is investigated, and the time limit value is temporarily set. Hold (step S100).
Then, it is investigated whether or not there are any unprocessed requests remaining in the time-limited order queue 101 (step S101).

At this time, if there is no unprocessed request in the examined queue, the request is registered in the time-limited order queue 101 as it is. If there are unprocessed requests remaining, is there a same value for the time limit value of the registration request temporarily held and the time limit value of the unprocessed requests remaining in the queue? Is investigated (steps S102 and S103).

When there is no request having the same time limit value in the queue, the registration request is inserted into the time limit order queue 101, and the process moves to the process of regenerating the queue list. When a request having the same time limit value is found, the priority order of the request and the registration request is investigated, and which request has a higher priority order (step S104,
S105).

When the priority of the registered request is higher than the request in the queue, the request is inserted into the queue immediately before the request having the same time limit value in the queue, and when the priority is low, it is immediately after (step S106, 1
07).

Next, from the time limit value of the registration request, the order in the time limit order queue 101 is examined, and the request is inserted into the corresponding queue (regeneration of the order in the queue). : Step S108). Subsequently, a request process for the access destination queue 102 inside the storage medium, which is another queue, is performed.

Here, the order of access destinations in the storage medium 105 will be described. In the following description, the case where the storage medium 105 is a disk has been described, but the storage medium 105 according to the present invention is not limited to the disk storage system, and includes all storage media that are sequentially accessed.

When the storage medium 105 is a disk, it has many tracks on its surface. The tracks are arranged concentrically from the center of the disc, and the track numbers are assigned to the tracks. Inside the track, there are areas divided into smaller units. This area is called a sector. The disk is accessed in this sector unit.

Access requests to the disk are "tracks 1 and 5 sectors", "tracks 5 and 2 sectors", "tracks 1 and 10 sectors", and "tracks 5 and 1".
When issued in the order of “3 sectors”, the heads for reading the sectors move in order of track 1 → track 5 → track 1 → track 5 because they are processed in the order issued in the conventional request method. It requires three head movements to fulfill the request. Since it takes a certain amount of time to move the head, the number of times of moving the head should be as small as possible in consideration of the processing time. In addition, when the head travels over a large amount of track, the head travel time increases, so it is preferable that the head travel time is also small.

In the request processing in the order of access destinations inside the storage medium 105 used in the present invention, in the case of a disk, accesses of the same track are continuous in ascending order (or descending order), and access of one track is 1 Even when accessing only a large number of tracks, requests should be issued so that the track numbers continue in ascending (or descending) order to minimize the number of head movements or the amount of head movement. Operate.

In the above-mentioned case, the present invention is the storage medium 1.
The requests issued in the order of access destinations inside 05 are "track 1, sector 5", "track 1, sector 10",
“Tracks 5, 2 sectors”, “tracks 5, 13 sectors”, the head is moved from track 1 to track 5, and the number of head movements is only once. For example, if the moving time of the head per movement is 10 msec, the process will be completed 20 msec earlier than the conventional method.

Next, referring to FIG. 4, as in the case of the time-limited forward queue 101, the value of the access destination among the parameters added to the registration request is investigated and the value of this access destination is temporarily retained. (Step S109).

Subsequently, it is investigated whether or not there are any unprocessed requests remaining in the access destination sequential queue 102 inside the storage medium (step S110).

At this time, if there is no unprocessed request in the examined queue, the request is directly registered in the access destination queue 102 inside the storage medium. If there are unprocessed requests remaining, is there an identical value for the access destination of the registration request temporarily held earlier and the access destination value of the unprocessed requests remaining in the queue? (Step S
111, S112).

If no request having the same access destination value exists in the queue, the registration request is inserted into the access destination sequential queue 102 inside the storage medium, and the queues are queued. Move on to the process of regenerating. When a request having the same access destination value is found, the priority order of the request and the registration request is checked, and which request has the higher priority order is checked (steps S113 and S114).

When the priority of the registered request is higher than that of the requests in the queue, the request insertion destination is immediately before the request having the same access destination value in the queue, and when the priority is low, it is immediately after. Yes (Step S
115, 116).

From the value of the access destination of the registration request, the order of the access destination order queue 102 inside the storage medium is examined, and the request is inserted into the queue in the corresponding order. (Regeneration of order in queue) (Step S
117).

The same requests are stored in the two queues, although the processing order is different. Therefore, if you check one of the queues and the request inserted in the queue has no registered request in the queue, and the inserted request is the first request in that queue, The request execution block 103 is notified of the request registration.
If it is not the first request, the process waits for registration of a new request (steps S118 and 119).

FIG. 5 shows the request execution block 10 of FIG.
6 is a flowchart for explaining the operation of the processing step of No. 3. In this embodiment, the queue generation block 1
Upon receiving the request registration notification from 00, the request execution block 103 starts operation. After the operation starts,
The request execution block 103 determines whether the queue currently fetching the request is the time-limited sequential queue 101 or the access destination sequential queue 1 inside the storage medium.
Whether it is 02 is checked by referring to a use queue flag (described later) (step S200).

Then, it is investigated whether or not there is a request in the queue currently being used, and if there is no request, the state of waiting for a request is entered (step S
201, 202).

Next, the parameters of the request located at the head of the queue are examined. If the expiration date is exceeded, the request request source is notified of the expiration date and this request is deleted from the queue. later,
The queue is checked again (steps S203, 20).
4,205).

If the request has not expired, the time limit parameter is subsequently checked. At this time, when the use queue flag (described later) is the time-limited order queue 101, only the first request is investigated, and when the access destination order queue 102 inside the storage medium, all requests in the queue are examined. To investigate the. That is, in the case of the access destination sequential queue 102 inside the storage medium, it is checked whether or not there is any request in the queue whose time limit is exceeded.

If the time limit of the request is exceeded, a flag is set in the time limit order queue 101, and if the time limit is not exceeded, a flag is set in the access destination order queue 102 inside the storage medium. (Steps S206, 207, 208). Since the queue currently used is determined at this point, the first request in the queue is executed (step S209).

After that, the request execution result is notified, the executed request is deleted from the time limit order queue 101 and the access destination order queue 102 inside the storage medium, and the processing order of the requests remaining in the queue is deleted. It is reconstructed (steps S210 and 211).

This completes the processing of one request. When the request processing is completed, it is checked again whether there are any requests remaining in the queue. If no request exists, a new request registration waiting state is entered, and if there is a request, the remaining requests are processed by the procedure described above.

As described above, according to the present embodiment, since the present invention is configured as described above, the following effects can be obtained. The first effect is that the disk processing and noise can be balanced even for stream data only. The reason is that the queue with the time constraint is given priority, and the queue with the access destination is given priority when the time constraint has a margin, and the queue with the access destination is given priority when the disk processing capacity has a margin. As a result, the moving distance of the head is shortened and noise can be reduced.

The second effect is that even if stream data and non-stream data are mixed, the demand for the disk system can be realized to the maximum extent.

A third effect is that the non-stream data is generally time-constrained less than the stream data, and the time-constraint processing is prioritized, so that the time-constraint can be performed even when non-stream data is mixed. This means that it is possible to minimize the effect on strong stream data. Non-stream data generally has less time constraints than stream data. Since the present invention gives priority to the time constraint processing, even when non-stream data is mixed, it is possible to minimize the influence on stream data having a strong time constraint.

(Second Embodiment) Although the storage medium scheduler can be realized by software on the system, the queue generation block 1 which is a constituent element
00, the request execution block 103, the time limit order queue 101, and the access destination order queue 102 inside the storage medium can be realized by hardware.

For example, the two queues are composed of dedicated memories, the functions of the queue generation block 100 and the request execution block 103 are realized by using a dedicated LSI (ASIC) for a specific customer, and Dedicated LS
With the configuration in which the storage medium controller 104 and the storage medium 105 are connected to the I (ASIC), the function of the present invention can be realized.

Further, a queue generation block 100, a request execution block 103, a time limit order queue 101
By implementing the functions of the storage medium 105 together with the functions of the access-destination queue 102 inside the storage medium together with the functions of the storage medium 105 (implementation into ASIC), all the functions of the present invention can be realized by one piece of hardware. Become.

It should be noted that the present invention is not limited to the above-described embodiments, and it is apparent that the embodiments can be modified appropriately within the scope of the technical idea of the present invention. Further, the number, positions, shapes, etc. of the above-mentioned constituent members are not limited to those in the above-mentioned embodiment, and the number, positions, shapes, etc. suitable for carrying out the present invention can be adopted. Moreover, in each figure, the same components are denoted by the same reference numerals.

[0072]

Since the present invention is configured as described above, it has the following effects. The first effect is that the disk processing and noise can be balanced even for stream data only. The reason is that the queue with the time constraint is given priority, and the queue with the access destination is given priority when the time constraint has a margin, and the queue with the access destination is given priority when the disk processing capacity has a margin. As a result, the moving distance of the head is shortened and noise can be reduced.

The second effect is that even if stream data and non-stream data are mixed, the demand for the disk system can be realized to the maximum extent.

The third effect is that the non-stream data is generally less time-constrained than the stream data, so that the time-constraint processing is prioritized. This means that it is possible to minimize the effect on strong stream data. Non-stream data generally has less time constraints than stream data. Since the present invention gives priority to the time constraint processing, even when non-stream data is mixed, it is possible to minimize the influence on stream data having a strong time constraint.

[Brief description of drawings]

FIG. 1 is a configuration diagram of a storage medium scheduler for explaining a first embodiment of the present invention.

FIG. 2 is an example of request order generation in a queue.

FIG. 3 is a flowchart for explaining the operation of the first half of the processing steps of the queue generation block of FIG.

4 is a flow chart for explaining the operation of the latter half of the processing steps of the queue generation block of FIG.

5 is a flowchart for explaining the operation of processing steps of the request execution block of FIG. 1. FIG.

[Explanation of symbols] 100 ... Queue generation block 101 ... Time-limited order queue 102 ... Queue in order of access destination inside storage medium 103 ... Request execution block 104 ... Storage medium controller 105 ... Storage medium 110 ... Inside the queue created earlier 111 ... Registered request 112 ... Inside the queue after sequence regeneration

─────────────────────────────────────────────────── ─── Continued front page (58) Fields surveyed (Int.Cl. ⁷ , DB name) G06F 3/06-3/08 G06F 12/00 G11B 20/10 G06F 1/00 G11B 27/10 H04N 5 / 76

Claims (12)

(57) [Claims] 1. A storage medium scheduler for optimally processing a request for a storage medium , comprising a priority, a time limit, an access destination inside the storage medium, and
In the request with a parameter that specifies the expiration date
On the other hand, optimization of access destination inside the storage medium is prioritized
Means for generating a reserved first queue and said time limit
Means for generating a second queue optimized in sequence;
The above priority, so that the processing of the storage medium is optimized.
Time limit, parameter of access destination inside the storage medium
To plan the execution order for the above requests based on
The first queue and the second queue
Scan and execute the most suitable request and wait for the first wait
Means for updating the queue and the second queue, and
A means for transmitting the request execution result to the request request source
The request for the storage medium in the storage medium
A queue prioritized for optimization of the access destination of
Two types of queues, which are queues optimized according to the time limit order
And refer to the two types of queues to store the
A queue with priority given to optimization of access destinations inside the medium
Scan and request if anything exceeds the time limit
Queues whose strike order is optimized according to the time limit order
Process the requests in the order of
Optimization occurs in the storage medium inside the storage medium
Can process requests in priority queue order.
To switch the request execution order in real time with
A storage medium scheduler having a configuration. 2. The storage medium scheduler according to claim 1 , wherein the storage medium includes video data for multimedia applications including video on demand. 3. The storage medium scheduler according to claim 1, wherein the storage medium includes audio data for multimedia use including video-on-demand. 4. The storage medium scheduler according to claim 1, wherein the storage medium includes data other than video and audio for multimedia use including video-on-demand. 5. The storage medium scheduler according to claim 1 , further comprising means for instructing the priority according to a type of data recorded in the storage medium. 6. The storage medium scheduler according to claim 1, wherein the storage medium is a magnetic storage disk. 7. The storage medium scheduler according to claim 1, wherein the storage medium is an optical storage disk. 8. The storage medium scheduler according to claim 1, wherein the storage medium is a magnetic tape. 9. A request for a storage medium is optimally screened.
Rear with scheduling means for scheduling
In the real-time system, the scheduling means prioritizes optimization of the access destination inside the storage medium with respect to the request including parameters for specifying priority, time limit, access destination inside storage medium, and expiration date. a queue that is, constitute the optimized queue by the time limit order, in the request is typically that was being processed by the order of the storage medium inside the access destination optimize prioritized queue, When the request that exceeds the time limit is about to be processed, the processing order is set to the time system.
The request is processed by changing to a queue optimized in a limited order, and if all the requests in both queues do not exceed the time limit, optimization of the access destination inside the storage medium is performed again. Are scheduled to be processed in priority queue order .
Real-time system that. 10. The scheduling means, when a new request is generated, based on the parameters of the request, the queue prioritized optimization of an access destination in the storage medium, and the time limit order.
A real-time system according to claim 9 , characterized in that it comprises means for reconstructing the content for an optimized queue. 11. The real-time system according to claim 10 , comprising two queues, and having means for switching the processing order in real time according to the state of the request to be processed. 12. A scheduling method for scheduling execution of a request to the storage medium, which has parameters for specifying a priority, a time limit, an access destination inside the storage medium, and an expiration date, the method being added to the request. Based on the parameters
The first priority is given to the optimization of the access destination inside the storage medium.
In addition to configuring the queue of 1,
Constructing an optimized second queue and said first waiting when a new request is requested
A process for reconfiguring the queue and the second queue.
And the priority of optimization of the access destination inside the storage medium.
The time limit existing in the queue is adjusted when processing the request.
And a step existing in the queue optimized in the time limit order.
Optimizing the process of checking the time limit during request processing and the access destination inside the storage medium to be executed
If there is a previous request in the queue
If the time limit is exceeded, it is optimal according to the time limit order
Process of transferring execution order to optimized queue, and waiting line optimized in the time limit order to be executed
If the request in the column is within the time limit
In this case, optimization of the access destination inside the storage medium is favored.
The step of shifting the execution order to the queue that was preceded, and the request to be executed exceeds the expiration date.
If there is a request, the process of immediately stopping the processing of the request
Scheduling method and having a degree.