JP2001101058A - Remote access method to secondary storage device - Google Patents

Abstract

PROBLEM TO BE SOLVED: To shorten the time of read/write processing to a remote secondary storage device through an inter-computer connecting network. SOLUTION: A library 11 linked with a client program 5 divides read/write calling from the client program 5 to a remote secondary storage device 4 into a plurality of read/write requests, and transmits them through an inter-computer connecting terminal 3 to a sever computer 2 in a batch at once. A file server 10 of the sever computer 2 receives the read/write requests, and accesses the requested data on the secondary storage device 4 and transmits one response to the client program being the origin of the request in response to one received read/write request. This processing is multiplexed and executed in response to the plurality of read/write requests so that the inter-computer communication time can be overlapped on the physical read/write processing time to the secondary storage device 4 and the read/write processing time to the remote secondary storage device is shortened.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a remote access method to a secondary storage device, and more particularly, to an application program running on a certain computer in an environment where a plurality of computers are connected by an inter-computer connection network. And a method for accessing data on a secondary storage device connected to another computer.

[0002]

2. Description of the Related Art There are several methods for accessing a secondary storage device (remote secondary storage device) connected to another computer.

[0003] One method is to use a Network File System (NF) as shown in FIG.
S). Data access in NFS is realized by software provided with a communication protocol layer called RPC (refer to the reference document “NFS & NIS”).
ASCII Publishing Bureau).

In the NFS, a read / write call from a client program is converted into an RPC request and transmitted to a process called an NFS daemon operating on a server computer using the RPC protocol. For example, in the case of a write call, write data is transmitted together with a request, but this data is fetched by the NFS daemon, written to the secondary storage device, and a response is transferred to the client program. N for read call
The FS daemon reads the target data from the secondary storage device and returns the data as a response to the RPC request.

When a read / write call from a client program is converted into an RPC request, it is broken down into a plurality of RPC requests as a read / write request for fixed-size data. Normally, the size of data accessed by one RPC request is 8 KB. The access requests are sent one by one to the NFS daemon, and after waiting for a response, the next RPC request is sent. That is,
Since the NFS daemon processes only one RPC request at a time, it is only necessary to secure an area for taking in 8 Kbytes of data. However, since one NFS daemon is occupied for one client program,
In order to process requests from a plurality of client programs, it is necessary to start a plurality of NFS daemons on the size calculator.

Another method is disclosed in, for example, Japanese Patent Application Laid-Open No. 4-929.
No. 43 discloses a high-speed LAN disk server. In this method, as shown in FIG. 17 (B), an environment including a disk server for managing a secondary storage device (limited to a disk) and a plurality of workstations, wherein a high-speed LAN is used as an inter-computer connection network. Is assumed. The disk server that manages the disk is dedicated hardware, and has a structure capable of simultaneously processing a plurality of read / write requests.

[0007] A client program runs on the workstation, and transmits a read / write request to the remote disk from the client program to the disk server by communication. For example, for a write request,
Send data with the request. On the other hand, the disk server temporarily stores the received data in the data buffer, writes the data on the disk, and notifies the requesting workstation by communication of the result. If the request is a read request, the disk server checks whether there is target data in the data buffer. If not, the data is read from the disk and stored in the data buffer. Further, the stored data is transmitted to the requesting workstation by communication.

In this method, a data buffer is provided in the disk server in order to relay data between the computer connection network and the secondary storage device. The data buffer is used by dividing an area for each workstation. Lead /
Write requests are made in fixed length units called frames.

In each of the two methods described above, a read / write call from a client program is decomposed as a fragmentary read / write request to a plurality of fixed-length data, and a part for managing a secondary storage device is used. (Server computer, disk server) to enable processing of read / write calls to the remote secondary storage device.

[0010]

However, since each fragmented fragment read / write request is transmitted one by one, and after the processing result is transmitted, the next fragment read / write request is transmitted, communication is performed many times. Will do. One lead /
The flow of the write call process is as shown in FIG. 17C, where (fragment read / write request transmission) → (secondary storage device read / write) → (response reply) → (fragment read / write request transmission) ) →..., And after all fragment read / write requests are completed, the result of the read / write call of the client program is returned. Therefore, the time required for remote secondary storage device access from the viewpoint of the client program appears to require extra round-trip communication time corresponding to the number of fragment read / write requests.

Therefore, according to the present invention, most of the communication time in the data access time to the remote secondary storage device is made invisible to the client program, and it is as if the secondary storage device is connected to the computer running the client program. The aim is to achieve such access times.

[0012]

According to the present invention, in order to achieve the above object, a first computer and a client program running on a first computer can communicate with each other through the first computer and an inter-computer connection network. In the method for accessing data on the secondary storage device connected to the second computer, the first computer and the second computer use the following (1)
Any one of the processes from (4) to (4) is executed.

(1) The first computer divides a read / write call by a client program into read / write requests to data of a fixed size, and divides the read / write requests obtained by the division all at once. After transmitting the read / write request to the second computer and confirming that a response has been returned from the second computer for all of the transmitted read / write requests, the read / write request is transmitted to the client program.
Returns the result of a write call. On the other hand, the second calculator is
A read / write request from the first computer is received, and in response to the received read / write request, data on the requested secondary storage device is accessed, and then a response is sent to the requesting client program. The process of transmitting is multiplexed and executed for a plurality of read / write requests.

As described above, the read / write call by the client program is divided into read / write requests for data of a fixed size and transmitted to the second computer all at once. By multiplexing the requests, the physical read / write processing of the secondary storage device in the second computer and the communication processing in the inter-computer connection network are performed in parallel. This overlaps with the read / write processing time, and the overall processing time is reduced.

(2) The first computer divides the read / write call by the client program into read / write requests for data of a fixed size, and the number of read / write requests obtained by the division is determined in advance. When the threshold value is equal to or less than the threshold value, the data is transmitted to the second computer all at once. When the threshold value is exceeded, a read / write request for the threshold value is first transmitted to the second computer. Whenever one response from the second computer corresponding to the read / write request is confirmed, one remaining read / write request is transmitted to the second computer, and all of the transmitted read / write requests are transmitted. After confirming that a response has been returned from the second computer, the result of the read / write call is returned to the client program. The second computer executes the same processing as in the above (1).

As described above, the threshold value is set as the number of requests that can be received at one time by the second computer, and the number of read / write requests transmitted first is suppressed to be equal to or less than the threshold value. Can be suppressed below a certain value. In addition, 1
Since one read / write request is transmitted for one response, the number of read / write requests received by the second computer is always equal to or less than the threshold value, and the load on the second computer is continuously suppressed to a certain value or less. be able to.

(3) The first computer divides the read / write call by the client program into read / write requests for data of a fixed size, and transmits the read / write request obtained by the division. Transmitting a reservation request to the second computer, determining a threshold value from the number of reservations included in a response from the second computer to the transmitted reservation request, and obtaining the divided value. When the number of read / write requests is equal to or less than the threshold value, all of the requests are transmitted to the second computer at once, and when the number of read / write requests exceeds the threshold value, the read / write requests for the threshold value are first sent to the second computer. , And each time a response from the second computer corresponding to a read / write request is confirmed, one remaining read / write request is After confirming that responses have been returned from the second computer for all of the transmitted read / write requests, the computer requests the second computer to cancel reservations, and Returns the result of the read / write call to the client program. On the other hand, in response to a reservation request from the first computer, the second computer sets the number of read / write requests permitted for one client program according to the current load status of the own computer as the reservation number. Notifying the first computer and securing a memory amount commensurate with the number of reservations, receiving a read / write request from the first computer, and responding to one received read / write request to the requested secondary storage device After accessing the above data, a process of transmitting one response to the requesting client program is multiplexed and executed for a plurality of read / write requests.
Release the reserved memory amount in response to a reservation release request from the computer.

As described above, the number of reservations permitted for one client program is notified to the first computer in accordance with the dynamically changing load state of the second computer. As a threshold value,
By making a read / write request in accordance with this threshold value, it is possible to control the load generated in the second computer so as not to exceed a certain value. Therefore, even if the second computer deals with a plurality of client programs, it does not exceed the processing capacity of the second computer.

(4) The first computer divides the read / write call by the client program into read / write requests for data of a fixed size, and one of the read / write requests obtained by the division. A reservation + read / write batch request in which a reservation request is added to one read / write request is transmitted to the second computer, and a reservation from the second computer for the transmitted reservation + read / write batch request is transmitted. A threshold value is determined from the number of reservations included in the response, and when the remaining number of read / write requests obtained by the division is equal to or smaller than the threshold value, 1 is set.
A reservation release request is added to one read / write request to make a reservation release + read / write batch request, and then all are sent to the second computer at once, and when the threshold value is exceeded, the threshold is first set. After transmitting the read / write request for the value to the second computer, each time a response from the second computer corresponding to the read / write request is confirmed, one remaining read / write request is transmitted to the second computer. The request is transmitted to the second computer and the last one read / write request is transmitted as a reservation release + read / write batch request to which a reservation release request is added, and the second computer is transmitted for all the transmitted requests. After confirming that a response has been returned from the client program, read /
Returns the result of a write call. On the other hand, the second calculator is
A request is received from the first computer, and if the received request is a reservation + read / write batch request, the number of read / write requests allowed for one client program according to the current load status of the own computer is the number of reservations. And notifies the first computer and secures a memory amount commensurate with the number of reservations, then accesses the requested data on the secondary storage device and transmits one response to the client program of the request source. When the received request is a read / write request, the process of accessing the requested data on the secondary storage device and transmitting one response to the requesting client program is performed. Is a reservation release + batch read / write request, after accessing the requested data on the secondary storage device, The process of the read / write request being processed generated from program releases the amount of memory to the reserved upon completion all, with respect to multiple requests, executes multiplexes.

As described above, by introducing the reservation + read / write batch request that requests the reservation request and the read / write request collectively, the first computer does not send the reservation request as another request first. In addition, since the read / write processing is performed subsequent to the reservation response on the second computer side, the communication time of the reservation response can be overlapped with the physical read / write time. Also, if a subsequent read / write request is transmitted from the first computer during a physical read / write in the second computer, their communication time may overlap the physical read / write time. it can. Further, by introducing a reservation release + read / write batch request that requests the reservation release request and the read / write request collectively, the reservation release request does not have to be sent as a separate request at the end, and the communication corresponding to the request is performed. Time can be reduced.

[0021]

DESCRIPTION OF THE PREFERRED EMBODIMENTS Next, embodiments of the present invention will be described in detail with reference to the drawings.

Referring to FIG. 1, an example of a computer system to which the present invention is applied includes a computer 1 on which a client program operates, a server computer 2, an inter-computer connection network 3 for connecting these computers to each other, and And a secondary storage device 4 connected to the server computer 2.

The file server 10 on the server computer 2
A process called is running. File server 10
Receives a read / write request by communication, reads / writes data on the secondary storage device 4 connected to the server computer 2 according to the request, and returns a response by communication. In the example of FIG. 1, the functions that the server computer should have are realized by operating a process called a file server 10 on a general-purpose computer, but other similar functions are implemented in the OS of the general-purpose computer. An embodiment or an embodiment realized by adding firmware to a dedicated computer is also conceivable.

In the computer 1 on which the client program operates, a client program 5, which is an application program, operates. In the example of FIG. 1, a dedicated program called a library 11 is linked to the client program 5 in order to transmit the read / write call from the client program 5 to the server computer 2. The library 11 generates an appropriate read / write request in accordance with a read / write call from the client program 5 and transmits the request to the file server 10. Request / response exchanges between the library 11 and the file server 10 follow a transfer method described later, and transmission of requests is controlled by the library 11. Note that, in place of the library 11, an embodiment in which a similar function is implemented in an OS of a general-purpose computer is also conceivable.

Some embodiments will be described below, but before that, the file server 1 which is the premise of them will be described.
A basic concept for multiplexing a plurality of requests with 0 will be described.

When attention is paid to one request, a typical processing sequence is, as shown in FIG. 2A, composed of processing of a series of actions with some interruptions. That is, a corresponding initial action is determined according to the type of the request, and after returning to the computer 1 on which the client program 5 operates in the final action through several intermediate actions as the subsequent processing proceeds. The processing of one request is completed. Therefore, by processing the action of another request at the interruption point, multiple processing of the request can be realized.

The file server 10 manages the above-mentioned actions using queues (action queues), and executes a main loop as shown in FIG.
The file server 10 has a port (reception port) for receiving the request, and receives the request by reading it (S1). Next, an initial action corresponding to the received request is registered at the end of the action queue (S
2). Next, one action is taken from the head of the action queue (S3). Then, the action is processed (S4), and the process returns to step S1. By such a main loop, the file server 10 alternately processes the actions of a plurality of requests, and realizes multiplex processing of requests.

One of the interruption points of the request processing is a physical I / O.
O is issued. At this time, multiplex processing is realized by issuing an I / O asynchronously, ending the action processing, and processing the action of another request. Asynchronous I
When the / O is completed, an asynchronous I / O completion handler (not shown) operates, and the asynchronous I / O completion handler executes the processing shown in FIG. 2C.

First, the action following the action that issued the I / O is determined. There are several conceivable methods for this determination as follows. 1) A method of directly determining an action by designating an action as an argument of the handler. 2) A method in which an action subsequent to the action that issued the I / O is registered in a queue, and the handler determines the corresponding action by searching the queue.

Next, the determined action is registered at the head of the action queue. This registered action is
Thereafter, it is taken out and processed in the main loop of the file server 10.

The asynchronous I / O completion handler is started by interrupting at an arbitrary timing during the operation of the file server 10.
0 and the asynchronous I / O completion handler are mutually exclusive so as not to operate simultaneously.

Hereinafter, a method for accessing data on the secondary storage device 4 connected to the server computer 2 from the client program 5 operating on the computer 1 will be described with reference to some embodiments.

Embodiment 1 In this embodiment, when a read / write call is issued from the client program 5, the library 11 executes the processing shown in FIG.

First, a read / write call from the client program 5 is analyzed and divided into read / write requests for data of a certain size (L11). For example, when a certain fixed size is set to 8 KB, a read / write call from the client program 5 becomes
For example, if it is read or write of 24 KB data, 8 KB from the beginning of the requested data
read / write request for the next 8 KB and the next 8 KBy
read / write request for te and the last 8 KBy
It is divided into read / write requests for te.

Next, all the read / write requests obtained by division are transmitted all at once to the file server 10 of the server computer 2 via the inter-computer connection network 3 (L12).
Then, it waits for a response returned from the server computer 2 (L13). Waiting for a response is processing for reading the response receiving port of the client program 5, and if a response has not been received, it waits until a response is received.
When a plurality of read / write requests are transmitted, a plurality of responses are returned.

Next, when a response from the server computer 2 is received, the number of responses returned so far is checked to determine whether all responses have been returned (L14). If the number of responses is smaller than the number of divisions of the read / write request, that is, if all responses have not been returned, the process returns to step L13 and waits for a response. If all responses have been returned, the process returns to the call of the client program 5.

On the other hand, the file server 10 receives a read / write request from the library 11, and accesses the requested data on the secondary storage device 4 for one received read / write request, The process of transmitting one response to a plurality of read / write requests is multiplexed and executed. Since the method of multiplexing and executing requests has been described above, a sequence focusing on one read / write request will be described with reference to FIG.

First, the file server 10 secures a memory (data buffer) serving as a relay for temporarily storing read / write data according to the read / write request (S11). Next, the physical read /
A write is issued asynchronously, and its completion is waited for (S12).
Here, the first action ends, and the request processing is interrupted. Next, when the physical read / write is completed, a read / write response is transmitted to the request source (S13). This is the final action of the read / write request.

As described above, in the present embodiment, the read / write call by the client program 5 is divided into read / write requests for data of a fixed size and all are transmitted to the server computer 2 all at once. The server computer 2 performs multiplex processing of the read / write request of the secondary storage device, and the physical read / write processing of the secondary storage device in the server computer 2 and the communication processing in the inter-computer connection network 3 are performed in parallel. become. As a result, the communication time overlaps with the physical read / write processing time, and the overall processing time is shortened.

The manner in which the communication processing time and the physical read / write processing time overlap will be described with reference to FIG. FIG. 5 shows a case where a read / write call from the client program 5 is divided into four read / write requests and requested. First, the library 11 first sends four read / write requests to the server computer 2 at a time. Then, the file server 10 of the server computer 2 multiplexes the four read / write requests, but the physical read / write starts when the first request is received, so that the communication time of the subsequent request is the first read. / Overlaps with write processing time. As a result, most of the requested communication time is invisible from the client program 5. Further, at the end of each request, a response is returned to the library 11, while the file server 10 of the server computer 2 performs the physical read / write of the next request while processing.
As a result, most of the communication time of the response becomes invisible from the client program 5.

FIG. 5 shows that most of the communication time overlaps with the physical read / write time in the server computer 2. As a result, most of the communication time becomes invisible from the client program 5, and the read /
The processing time of the write call is reduced. However, the communication time 24 in FIG. 5 is not visible from the client program 5, but the communication time 23 of the first request and the last response is visible to the client program 5. However, as the amount of data to be accessed increases and the number of divisions increases, the time 23 becomes relatively smaller than the sum of the physical read / write times 22, and the read / write call processing time 2
1 is equivalent to the sum of the physical read / write times 22.

Further, responses are returned one after another from the server computer 2 to the computer 1 on which the client program 5 runs.
On the other hand, the library 11 knows that the processing of all the requests has been completed, so that the completion of the read / write call from the client program 5 can be determined.

Embodiment 2 In this embodiment, when a read / write call is made from the client program 5, the library 11 executes the processing shown in FIG.

First, as in the first embodiment, a read / write call from the client program 5 is analyzed and divided into read / write requests for data of a certain fixed size (L21).

Next, it is checked whether or not the number of divisions is larger than a predetermined threshold value (L22). And if the number of divisions is not greater than the threshold value,
All the read / write requests obtained by the division are transmitted all at once to the file server 10 of the server computer 2 via the inter-computer connection network 3 (L24). If the division number is larger than the threshold value, the division is performed. Of the read / write requests obtained as described above, only the number of read / write requests equal to the threshold value is transmitted at a time to the file server 10 of the server computer 2 via the inter-computer connection network 3 (L2
3). At this time, a read / write request that has not been transmitted is held internally. Then, it waits for a response returned from the server computer 2 (L25).

Next, when a response from the server computer 2 is received, it is confirmed whether or not a read / write request which has not been transmitted remains (L26). If there are read / write requests that have not been transmitted yet,
One read / write request is transmitted to the file server 10 of the server computer 2, and the process shifts to waiting for a response in step L25.
If it is determined in step L26 that no untransmitted read / write requests remain, the number of responses returned so far is checked to determine whether all responses have been returned (L28). If the number of responses is smaller than the number of divisions of the read / write request, that is, if all responses have not been returned, the process returns to step L25 and waits for a response. If all responses have been returned, the process returns to the call of the client program 5.

The processing of the file server 10 is the same as that of the first embodiment. A read / write request is received from the library 11, and one received read / write request is added to the requested data on the secondary storage device 4. After the access, a process of transmitting one response to the request source is multiplexed and executed for a plurality of read / write requests.

As in the first embodiment, the basic effects of the present embodiment include an effect that most of the communication time is made invisible to a client program by overlapping with physical read / write. However, in the first embodiment, all the read / write requests obtained by division are transmitted all at once.
When the data amount of the read / write call from the client program 5 is large, the number of read / write requests transmitted at one time increases, and the load on the server computer 2 and the inter-computer communication network 3 becomes excessive. If the load on the server computer 2 or the like becomes abnormally high, some of the read / write requests may disappear and not be processed, so that the read / write call from the client program 5 may fail.

On the other hand, in the present embodiment, the threshold value is set as the number of requests that can be received at one time by the server computer 2, and the number of read / write requests transmitted first is suppressed below the threshold value. The initial load on the server computer 2 can be kept below a certain value. Further, since one read / write request is transmitted for one response, the number of read / write requests received by the server computer 2 is always equal to or less than the threshold value, and the load on the server computer 2 is continuously reduced to a certain value or less. Can be suppressed.

FIG. 7 shows an example of a request / response exchange occurring between the server computer 2 and the client program 5 in this embodiment. In this example, the read / write call from the client program 5 is divided into eight read / write requests, and the threshold value is 4
Is set to

First, since the threshold value is 4, the library 11 first transmits only four read / write requests to the server computer 2. As a result, the number of read / write requests received first by the server computer 2 can be suppressed to four or less.

Thereafter, the file server 10 of the server computer 2 multiplexes the individual read / write requests and returns a response one by one. On the other hand, the library 11 transmits one read / write request to the server computer 2 every time one response is received. As a result, the number of read / write requests multiplexed by the server computer 2 is always four or less.

Embodiment 3 In this embodiment, when a read / write call is issued from the client program 5, the library 11 executes the processing shown in FIG.

First, as in the first embodiment, a read / write call from the client program 5 is analyzed and divided into read / write requests for data of a certain fixed size (L31).

Next, the reservation request specifying the number of divisions is transmitted to the file server 10 of the server computer 2 via the inter-computer connection network 3 and the response is waited for (L32).

Next, the file server 1 of the server computer 2
When a response to the above-mentioned reservation request is received from 0, the number of reservations included in the response is extracted and set as a threshold value (L33).

Next, it is checked whether or not the number of divisions at step L31 is larger than the threshold value (L3).
4). If the number of divisions is not larger than the threshold value, all read / write requests obtained by division are all sent to the server computer 2 via the inter-computer connection network 3 at once.
(L36), and when the number of divisions is larger than the threshold value, only the number of read / write requests equal to the threshold value among the read / write requests obtained by division at one time between the computers. Connection network 3
Is transmitted to the file server 10 of the server computer 2 via (L35). At this time, a read / write request that has not been transmitted is held internally. Then, it waits for a response returned from the server computer 2 (L37).

Next, when a response from the server computer 2 is received, it is confirmed whether or not a read / write request which has not been transmitted remains (L38). If there are read / write requests that have not been transmitted yet,
One read / write request is transmitted to the file server 10 of the server computer 2 (L39), and the flow shifts to a response waiting state in step L37. If it is determined in step L38 that no untransmitted read / write request remains, the number of responses returned so far is checked to determine whether all responses have been returned (L3a). If the number of responses is smaller than the number of divisions of the read / write request, that is, if all responses have not been returned, the process returns to step L37 and waits for a response.

If all responses have been returned, a request to cancel the reservation in the server computer 2 (reservation release request) is transmitted to the file server 10 of the server computer 2 through the inter-computer connection network 3 and Wait for a response (L3b).
Then, after a response to the reservation release request is returned, the process returns to the call of the client program 5.

In the case of this embodiment, the file server 10
In addition to the read / write request processing, reservation request processing and reservation release request processing are performed. Since the read / write request processing is almost the same as in the first and second embodiments, the reservation request processing and the reservation release request processing will be described below.

FIG. 9 shows a sequence focusing on one reservation request process. When the reservation request is transmitted from the library 11 as described above, the file server 10
Calculates the number of read / write requests allowable for the client program (S31).

This allowable number is determined by the current load status of the file server 10. Specifically, it is substantially determined by the current remaining memory capacity that the file server 10 can use for processing the read / write request. For example, if the total amount of memory that can be used by the file server 10 to process a read / write request is A and the amount of memory required to process one read / write request is B, [A / B]
Is the processing load upper limit of the file server 10. Further, if the reservation of C read / write requests is currently accepted, the total of the current allowable number (total allowable number) is C,
The allowable number of read / write requests at this time is the processing load upper limit-C. Therefore, the file server 10 obtains, as the allowable number for the client program, a value obtained by subtracting the total of the current allowable number (total allowable number) from the number serving as the upper limit of the processing load.

Next, the smaller of the allowable number obtained in step S31 and the number of divisions specified in the reservation request is determined as the number of reservations (S32). Then, the total of the current allowable number is increased by the current reservation number (S33), and a response specifying the reservation number obtained in step S32 is transmitted to the request source (S34). By increasing the total of the current allowable number by the number of reservations this time, the amount of memory corresponding to the number of reservations this time can be read / generated by other client programs.
It can be prevented from being used in write request processing.
Therefore, it is possible to prevent a shortage of memory when securing a relay memory for temporarily storing read / write data according to a read / write request from the client program.

An embodiment in which the processing content of step S32 is changed to the number of reservations = min (allowable number, division number, maximum allowable number given per one client program) is also conceivable. In this way, the maximum allowable number given to one client program can be limited, and it is possible to prevent one client program from occupying the current allowable number.

The reservation request processing as described above is composed of one action.

FIG. 10 shows a sequence focusing on the processing of one reservation release request. As described above, when the reservation release request is transmitted from the library 11, the file server 10 updates the total of the current allowable number by subtracting the reservation number given to the client program (S35). As a result, the reserved memory amount is released, and the memory amount available for processing read / write requests from other client programs increases accordingly. Then, the end of the reservation release request processing is notified to the library 11 by a response (S36). Such reservation cancellation request processing is composed of one action.

When the threshold value is determined in advance according to the processing capacity of the server computer 2 as in the second embodiment, if the number of read / write requests exceeds the processing capacity of the server computer 2 if there is only one client program. Never come. However, the read /
When there are a plurality of client programs for transmitting the write request, the processing capacity of the server computer 2 may be exceeded.

On the other hand, in the present embodiment, the number of reservations permitted for one client program is notified to the library 11 in accordance with the dynamically changing load status of the file server 10 of the server computer 2. Since the number of reservations is set as a threshold value, and thereafter, a read / write request is made in accordance with the threshold value, the load generated on the server computer 2 can be controlled so as not to exceed a certain value. In other words, even if the server computer 2 deals with a plurality of client programs, it does not exceed the processing capacity of the server computer 2. Further, when all of the read / write requests from the client program are completed, the reservation number is reset, so that the memory can be effectively used for a reservation request of the client program generated thereafter.

FIG. 11 shows an example of a request / response exchange occurring between the server computer 2 and the client program 5 in this embodiment. Except for the reservation processing (a) and the reservation release processing (b), the exchange is the same as in the second embodiment. In the example of FIG. 11, the read / write call from the client program 5 is divided into five read / write requests, but it is assumed that the number of reservations from the file server 10 is three. Therefore, the library 11 sets the threshold value to 3, and first sends three read / write requests to the server computer 2. Subsequent processing is the same as the processing procedure of the second embodiment, except that after all read / write requests have been processed, the reservation release processing (b) is performed.
The process returns to the read / write call from the client program 5.

(Embodiment 4) In this embodiment, when a read / write call is issued from the client program 5, the library 11 executes the processing shown in FIG.

First, as in the first embodiment, a read / write call from the client program 5 is analyzed and divided into read / write requests for data of a certain fixed size (L41).

Next, it is checked whether the number of divisions is greater than 1 (L42). If the number of divisions is one, the one read / write request is transmitted to the file server 10 of the server computer 2 via the inter-computer connection network 3 without making a reservation request, and the reception of a response is waited (L44).

On the other hand, when the number of divisions is greater than 1, a reservation + read / write batch request in which a reservation request is added to one of the read / write requests obtained by division is: The request is transmitted to the file server 10 of the server computer 2 as one request, and the reception of the reservation response is waited (L43). Waiting for reception of a reservation response is processing for reading the response reception port of the client program 5, and if a reservation response has not been received, it waits until it is received.

Next, when a response to the reservation is returned from the server computer 2, the number of reservations included in the response is extracted and set as a threshold value (L45).

Next, it is checked whether or not the number of divisions -1 at step L41 is larger than the threshold value (L4).
6). If the number of divisions -1 is not larger than the threshold value, all the remaining read /
The write requests are transmitted all at once to the file server 10 of the server computer 2 via the inter-computer connection network 3 (L4
8). However, for one of the read / write requests, a reservation release request is added to make a reservation release + read / write batch request, which is transmitted as one request. If the number of divisions -1 is larger than the threshold value, only the read / write requests of the number equal to the threshold value out of the remaining read / write requests obtained by division are transmitted through the inter-computer connection network 3 at a time. To the file server 10 of the server computer 2 (L47). At this time, a read / write request that has not been transmitted is held internally. Then, the read /
Wait for a write response (L49).

Next, when a read / write response is received from the server computer 2, it is checked whether or not a read / write request that has not been transmitted remains (L4a).
If there are remaining read / write requests that have not yet been transmitted, it is checked whether the number is 1 (L4c). If not, that is, if there are two, there are two or more read / write requests that have not been transmitted yet. If you have one,
A write request is transmitted to the file server 10 of the server computer 2 (L4d), and the flow shifts to a response waiting state in step L49. On the other hand, when one read / write request has not been transmitted yet, a reservation release request is added thereto, and one request of reservation release + read / write batch request is transmitted to the file server 10 of the server computer 2 ( L4e), Step L4
Move on to wait for the response of No. 9.

If it is determined in step L4a that no untransmitted read / write request remains, the number of responses returned so far is checked to determine whether all responses have been returned (L4b). If the number of responses is smaller than the number of divisions of the read / write request, that is, if all responses have not been returned, the process returns to step L49 and waits for a response. If all responses have been returned, the process returns to the call of the client program 5.

In the case of this embodiment, the file server 10
In addition to the read / write request processing, reservation + read / write batch request processing and reservation release + read / write batch request processing are multiplexed and executed.

FIG. 13 shows a sequence focusing on one read / write request. The file server 10
First, a memory serving as a relay for temporarily storing read / write data according to the read / write request is secured (S41). Next, physical read / write is performed asynchronously to the secondary storage device 4, and the completion is waited (S42).

Next, when the physical read / write is completed,
It is checked whether or not the processing of the reservation release + read / write batch request is interrupted by the request from the same client program 5 (S43). If interrupted, it is checked whether there is another read / write request being processed from the same client program 5 (S44). If yes, the interrupted reservation release + read / write batch request has been interrupted. Is registered in the request queue (S45), and a response to the read / write request for which the physical read / write has been completed is transmitted to the request source (S46). On the other hand, if the processing of the reservation release + read / write batch request is not interrupted or there is no such reservation release + read / write batch request,
Alternatively, even if there is an interrupted reservation release + read / write batch request, if there is a processing read / write batch request from the same client program, step S45 is skipped and a response to the read / write request is requested. The original is transmitted (S46).

FIG. 14 shows a sequence focusing on the processing of one reservation + read / write batch request. As described above, when the reservation + read / write batch request is transmitted from the library 11, the file server 10 first calculates the number of read / write requests allowable for the client program (S47). This allowable number is determined by the current load status of the file server 10. Specifically, it is calculated as a value obtained by subtracting the total of the current allowable number (total allowable number) from the number serving as the upper limit of the processing load of the file server 10.

Next, the smaller of the allowable number obtained in step S47 and the division number -1 specified in the reservation request is determined as the reservation number (S48). Then, the total of the current allowable number is increased by the current reservation number (S4).
9) A response specifying the number of reservations determined in step S48 is asynchronously transmitted to the request source (S4a).

An embodiment is also conceivable in which the content of the processing in step S48 is changed to the number of reservations = min (allowable number, division number−1, maximum allowable number given per client program). In this way, the maximum allowable number given to one client program can be limited, and it is possible to prevent one client program from occupying the current allowable number.

The processing so far is almost the same as the reservation request processing in the file server 10 of the third embodiment. However, in this embodiment, after the reservation response is transmitted asynchronously, the process proceeds to the read / write request processing, so that the processing time of the read / write request and the communication time of the reservation response overlap.

In the processing of the read / write request, first, a memory serving as a relay for temporarily storing the read / write data according to the read / write request is secured (S4).
b). Next, the physical read / write is asynchronously performed on the secondary storage device 4, and the completion is waited (S4c). When the physical read / write is completed, a read / write response is transmitted to the request source (S4d).

On the other hand, FIG. 15 shows one reservation release + read /
This shows a sequence focusing on a write batch request. As described above, when a reservation release + read / write batch request is transmitted from the library 11, first, a read / write request process is performed, and then a reservation release process is performed. Lead /
In the write request processing, a memory serving as a relay for temporarily storing the read / write data according to the read / write request is secured (S4e), the physical read / write is asynchronously performed on the secondary storage device 4, and the completion is completed. Wait (S4
f). When the physical read / write is completed, it is checked whether there is another read / write request being processed from the same client program 5 (S4g), and if there is, the processing is interrupted until there is no more read / write request being processed. . The suspension is performed by registering the request in a suspension queue. The request registered in the suspension queue is re-registered from the suspension queue to the request queue in step S45 of FIG.

When there is no other read / write request being processed, the reservation is released, and the number of reservations is subtracted from the current total allowable number and updated (S4h). Thereby, the memory amount reserved for the client program is released. Then, a read / write response including a reservation release response is transmitted to the request source (S4i).

According to the present embodiment, since the reservation + read / write batch request is introduced, the library 11 does not need to send the reservation request as a separate request first, and the file server 10 succeeds the reservation response. Since the read / write processing is performed, the communication time of the reservation response can be overlapped with the physical read / write time. Further, if a subsequent read / write request is transmitted from the client program during the physical read / write in the server computer, the communication time can be overlapped with the physical read / write time.

Further, since the reservation release + read / write batch request is introduced, the reservation release request does not have to be sent as a separate request at the end, and the communication time can be reduced accordingly.

As described above, the communication time for the first and last reservation / reservation cancellation can be reduced. FIG.
Shows an example of a request / response exchange occurring between the server computer 2 and the client program 5. The first reservation + read / write batch request (c) is sent to the server computer 2 as one request. The server computer 2 first performs a reservation process, returns a reservation response, and simultaneously reads /
Move on to write processing. At this time, the communication times of the physical read / write and the reservation response overlap.

The read / write request processing after the reservation response is received by the library 11 of the computer 1 on which the client program 5 operates is the same as in the third embodiment. However, the difference is that the last request is a reservation release + read / write batch request. The processing of the last request corresponds to the processing of reservation release + read / write (d) shown in FIG. FIG.
In the example of No. 6, read from client program 5 /
Although the write call is divided into five read / write requests, it is assumed that the number of reservations from the file server 10 is three. For this reason, the library 11 sets the threshold value to 3, and after receiving the reservation response,
One read / write request is sent to the server computer 2. Subsequent processing is the same as the processing procedure of the second and third embodiments, except that when the number of remaining read / write requests becomes one, a request that combines reservation release and read / write (reservation release +
Read / write batch request). On the server computer,
Read / write processing and reservation release processing are performed in response to this request, and returned as one response. Thereby, the last communication time can be reduced.

The embodiments of the present invention have been described above.
The present invention is not limited to the above embodiments, and various other additions and changes are possible. For example, in FIG. 1, the computer 1 on which one client program 5 operates is connected to one server computer 2 via the inter-computer communication network 3, but the computer on which one or more client programs 5 operate Can be applied to a computer system in which a plurality of computers are connected to the inter-computer communication network 3. Further, the example in which only one secondary storage device 4 is connected to the server computer 2 has been described, but a plurality of secondary storage devices may be connected.

[0093]

As described above, according to the present invention, the following effects can be obtained.

According to the present invention, most of the communication time at the time of data access to the remote secondary storage device is made invisible to the client program, and it is as if the secondary storage device is connected to the computer running the client program. Such access time can be achieved.

According to the present invention, even when a client program accesses a large amount of data,
That is, even when the number of divided read / write requests increases, the amount of read / write requested at a time can be suppressed, and the processing load on the second computer can be prevented from becoming excessive. .

Furthermore, according to the present invention, even when a plurality of client programs make read / write requests to one second computer, the read / write requests are made at once according to the load on the second computer. Can be suppressed, and the processing load of the second computer can be prevented from becoming excessive.

Further, according to the present invention, the first computer does not need to send a reservation request as a separate request first,
On the computer side, the read / write processing is performed subsequent to the reservation response.
Can overlap with the time of the light. Also, if a subsequent read / write request is transmitted from the first computer during a physical read / write in the second computer, their communication time may overlap the physical read / write time. it can. Furthermore, the first computer does not need to send the reservation release request as a separate request at the end, and the communication time can be reduced accordingly.

[Brief description of the drawings]

FIG. 1 is a block diagram illustrating an example of a computer system to which the present invention has been applied.

FIG. 2 is a flowchart illustrating an example of a process for multiplexing a plurality of requests in a file server.

FIG. 3 is a flowchart illustrating an example of library processing according to the first exemplary embodiment.

FIG. 4 is a flowchart illustrating an example of a process of the file server according to the first embodiment that focuses on one read / write request;

FIG. 5 is a timing chart showing exchange of requests / responses between a client program and a server computer in the first embodiment.

FIG. 6 is a flowchart illustrating an example of a library process according to the second embodiment.

FIG. 7 is a timing chart showing exchange of requests / responses between a client program and a server computer in a second embodiment.

FIG. 8 is a flowchart illustrating an example of processing of a library according to the third embodiment.

FIG. 9 is a flowchart illustrating an example of processing of the file server according to the third embodiment that focuses on one reservation request.

FIG. 10 is a flowchart illustrating an example of a process of the file server according to the third embodiment that focuses on one reservation release request;

FIG. 11 is a timing chart showing exchange of requests / responses between a client program and a server computer in a third embodiment.

FIG. 12 is a flowchart illustrating an example of processing of a library according to the fourth embodiment.

FIG. 13 is a flowchart illustrating an example of a process of the file server according to the fourth embodiment that focuses on one read / write request;

FIG. 14 is a flowchart illustrating an example of processing of the file server according to the fourth embodiment that focuses on one reservation + read / write batch request;

FIG. 15 is a flowchart illustrating an example of processing of the file server according to the fourth embodiment, which focuses on one reservation release + read / write batch request;

FIG. 16 is a timing chart showing exchange of requests / responses between a client program and a server computer in a fourth embodiment.

FIG. 17 is an explanatory diagram of a conventional access method to a remote secondary storage device.

[Explanation of symbols]

 DESCRIPTION OF SYMBOLS 1 ... Computer on which client program operates 2 ... Server computer 3 ... Connection network between computers 4 ... Secondary storage device 5 ... Client program 10 ... File server 11 ... Library

Claims (5)

[Claims] 1. A method according to claim 1, wherein a client program operating on a first computer transmits data on a secondary storage device connected to a second computer communicable with the first computer via an inter-computer network. Wherein the first computer divides a read / write call by the client program into a read / write request for data of a fixed size, and the read / write obtained by the division. Prior to the transmission of the request, the total number of the read / write requests is set as a division number, transmitted to the second computer while being included in the reservation request, and a response from the second computer to the transmitted reservation request. A threshold value is determined from the number of reservations included in the request, and when the number of read / write requests obtained by the division is equal to or less than the threshold value, the threshold value is determined all at once. Transmitted to the second computer, said first read / write request threshold value component when exceeding the threshold value second
After transmitting one response from the second computer corresponding to the read / write request after transmitting to the second computer, transmitting one remaining read / write request to the second computer, For all read / write requests
After confirming that a response has been returned from the second computer, the second computer requests the second computer to cancel reservation, and returns a read / write call result to the client program. In response to a reservation request from the first computer, a quotient obtained by dividing the total memory capacity of the own computer by a data size used in one read / write request is set as a processing load upper limit, and the processing load upper limit is The difference obtained by subtracting the total number of reserved reservations is determined as the allowable number, and the smaller of the allowable number and the number of divisions included in the reservation request from the first computer is determined as the number of reservations. Notifying the first computer of the number, securing a memory amount commensurate with the reserved number, receiving a read / write request from the first computer, and responding to one received read / write request. Sa Transmitting a single response to the requesting client program after accessing the data on the secondary storage device, and multiplexing and executing the plurality of read / write requests. In response to a request to cancel a reservation from the server, the reserved memory amount is released, and the multiplexing of processing in the second computer is performed by a series of processing for one request from reception of a request to return of a response. It is realized by executing the action of another request at the interruption point of the action, and the
A remote access method to a secondary storage device, wherein an action subsequent to an action that issued / O is preferentially processed. 2. In the second computer, a quotient obtained by dividing a total memory capacity of the own computer by a data size used in one read / write request in response to a reservation request from the first computer is processed. An upper limit, a difference obtained by subtracting the total number of reservations currently reserved from the processing load upper limit is set as an allowable number, the allowable number, the number of divisions included in the reservation request from the first computer, and one client. The minimum value among the maximum allowable numbers to be given per program is determined as the number of reservations, the number of reservations is notified to the first computer, and a memory amount corresponding to the number of reservations is secured. 2. The method for remotely accessing a secondary storage device according to claim 1, wherein: 3. From a client program operating on a first computer, data on a secondary storage device connected to a second computer communicable with the first computer via an inter-computer network. Wherein the first computer divides a read / write call by the client program into a read / write request for data of a fixed size, and the read / write obtained by the division. A request for reservation including the total number of read / write requests as a division number is added to one of the requests, and the request is transmitted to the second computer as a reservation + read / write batch request. A threshold value is determined from the number of reservations included in a response from the second computer to the transmitted reservation request, and the read / write obtained by the division is determined. When the remaining number of write requests is equal to or smaller than the threshold value, a reservation release request is added to one read / write request to make a reservation release + read / write batch request, and then the second request is issued all at once. When the threshold value is exceeded, a read / write request for the threshold value is first sent to the computer.
After transmitting to the second computer, each time one response from the second computer corresponding to the read / write request is confirmed, one remaining read / write request is transmitted to the second computer, In addition, the last one read / write request is transmitted as a reservation release + read / write batch request to which a reservation release request is added.
After confirming that a response has been returned from the second computer, the result of the read / write call is returned to the client program, and the second computer receives and receives the request from the first computer. When the request is a reservation + read / write batch request, the quotient obtained by dividing the total memory capacity of the own computer by the data size used in one read / write request is set as the processing load upper limit, and the current reservation is made from the processing load upper limit. The difference obtained by subtracting the total number of reserved reservations is determined as the allowable number, and the smaller value of the allowable number and the division number included in the reservation request from the first computer is determined as the reservation number, and the reservation number is determined. Notifying the first computer,
After securing a memory amount commensurate with the reservation number, the requested request is read / accessed by accessing data on the requested secondary storage device and transmitting one response to the requesting client program. In the case of a write request, the process of accessing the requested data in the secondary storage device and then transmitting one response to the requesting client program is performed. In some cases, after the requested data on the secondary storage device is accessed, the secured memory amount is released when all of the read / write requests being processed generated from the requesting client program are completed. Multiplexing and executing a plurality of requests, and the multiplexing of the processing in the second computer is one of processing from receiving a request to returning a response. Realized by executing an action for other requests with break point of a series of actions relating to the request, also natural, I against the two temporary storage devices
A remote access method to a secondary storage device, wherein an action subsequent to an action that issued / O is preferentially processed. 4. The second computer, wherein the request received from the first computer is a reservation + read /
In the case of a write batch request, the quotient obtained by dividing the total memory capacity of the own computer by the data size used in one read / write request is set as the processing load upper limit, and the total number of reservations currently reserved is calculated from the processing load upper limit. The reduced difference is regarded as an allowable number, the allowable number, the number of divisions included in the reservation request from the first computer, and the minimum value of the maximum allowable number given per client program as the number of reservations. Determine and notify the first computer of the number of reservations, secure a memory amount corresponding to the number of reservations, and then access the requested data on the secondary storage device to request the client program. 4. The method according to claim 3, wherein one response is transmitted to the secondary storage device. 5. The secondary storage device according to claim 1, wherein a plurality of said first computers are connected to said second computer through said inter-computer connection network. Remote access method.