JPS62287359A - Control system for simultaneous file access in loosely coupled multi-processor system - Google Patents

Abstract

PURPOSE:To improve the processing capacity of a system by being conscious of the hierarchies of a block and a record, and executing a shared exclusive control by a recording unit, in case of an access competition between tasks operated on the same computer. CONSTITUTION:With respect to a recording shared exclusive control means 11, a lock request is executed in a shared mode by designating a record, and with respect to a shared buffer managing means 21, a buffer securing request is executed by designating a block containing a request record. From a buffer obtained by the shared buffer managing means 21, the requested record is transferred to the record store area of a request task, and with respect to the shared buffer managing means 21, a buffer release request is executed in a reference mode by designating a block processed already.

Description

Detailed Description of the Invention 3. Detailed Description of the Invention [Field of Industrial Application] The present invention relates to a simultaneous access control method for computationally slow-open shared files in a loosely coupled multiprocessor system.

[Conventional technology]

In conventional file simultaneous access control in a loosely coupled multiprocessor system, the control unit is a block or a file, and control is not performed on a record-by-record basis.

On the other hand, in a single processor system or a tightly coupled multiprocessor/processor system, simultaneous file access control has been performed on a record-by-record basis, for example, using a configuration as shown in FIG. In this figure, shared exclusive control is a basic function used by simultaneous file access control, and controls the sharing or exclusive use of resources among multiple tasks using the requested logical resource name as a key.

In this control, since it is not concerned which physical resource a logical resource name corresponds to, it is possible to associate the logical resource name with a block or with a record. In addition, shared buffer management allocates a shared buffer area so that if the buffer corresponds to the same block, the same buffer is shared among multiple tasks requesting the block. Tasks that perform simultaneous access on a record basis request shared exclusive control by associating records as logical resource names, and file updates are assigned by shared buffer management in association with the block containing the record. This is done via a buffer area.

The necessity of updating files via the shared buffer in the above example will now be explained with reference to FIG. 3. This figure shows the operation when a buffer area is provided for each task in a single processor system. In order to update record R1, task T1 reads record 1 into the exclusive secured edge and block B1 containing record R1 into its own task's buffer area. Note that input/output operations for record box 1 alone are not possible.

Next, task T2 exclusively secures record box 2 in order to update record box 2. Record R1' and record R
2 is a separate source rounding, no waiting occurs. After that, the same block B1 containing record R2 is read into the buffer area of the invoking task. After task T1 changes record 1 to 1' in its own task's buffer area, it writes this changed block Bl, and then task T2
changes record R2 to 几2' and then writes this changed block B1. As a result, task T2 does not have this change from R1 to R1, so task T1
The update results performed in task 2 will be lost due to the overwriting of the update results performed in task 2.

[Problem that the invention seeks to solve]

In the above-mentioned conventional technology, in a loosely coupled multiprocessor system, even if it is desirable to perform simultaneous file access control on a record-by-record basis, simultaneous file access control must be performed on a block-by-block basis. As a result, the frequency of waiting times due to access conflicts increases, and the processing time of tasks that access files simultaneously increases, which is a problem.

The reason why it is difficult to control simultaneous file access on a record-by-record basis in a loosely coupled multiprocessor system, which is possible in a single processor system or a tightly coupled multiprocessor system, is that the input/output buffer area (usually main memory The reason is that the information (taken by the computer) cannot be shared between computers. Therefore, if the unit of shared exclusive control is simply set as a record, the result of updating the shared file cannot be guaranteed due to the same logic as when a buffer area is provided for each task in the single processor system described above.

[Means for solving problems]

The simultaneous file access control method in a loosely coupled multiprocessor system of the present invention provides simultaneous access control to a file on a secondary storage device shared by a plurality of computers making up the loosely coupled multiprocessor system. When the control unit has a block, which is an integral multiple of the physical input/output unit, and a record, which is a control unit smaller than the block, it is possible to operate on the same computer in response to file access requests in record units. If there is an access conflict between tasks, share exclusive control is performed in the requested record unit,
In addition, if there is an access conflict between tasks running on different computers, there is a record sharing exclusive control means that performs shared exclusive control in units of blocks containing requested records, and a shared buffer area that exists for each individual machine. If the request is for allocation of a buffer corresponding to the same block, the shared buffer management means controls the allocation so that the same area is shared among multiple tasks requesting the buffer. (When accessing a file in record units, shared exclusive control is performed by the record shared exclusive control means, and physical input/output operations are configured to be performed via a buffer area allocated by the shared buffer management means. .

〔Example〕

An embodiment of the present invention will be described with reference to the drawings.

Referring to FIG. 1, one embodiment of the present invention includes a shared exclusive control device 1 having a record shared exclusive control means 11, a shared buffer management means 21, and a shared buffer area 221.
* Loosely coupled multi-processing, a total of 3 that make up the t system
It consists of an I machine 2 and a secondary storage f3 in which files shared among multiple computers exist. The system of this embodiment is composed of two computers 2, and each computer 2
It has a shared exclusive control device 1 which is a device exclusively for shared exclusive control and is channel-connected to a shared exclusive control device.

Furthermore, a magnetic disk device is used as an example of the secondary storage device 3. In addition, either or both of the total N
Machine 2 (Z Ngle Processor) T-tightly coupled multiprocessor system! It is also possible to implement it by changing the @ configuration. - It is still possible to implement the shared exclusive control device 1 by incorporating it into one of the computers 2, or by implementing it by incorporating the shared exclusive control device 1 into one of the computers 2.
It is also possible to implement this in a manner in which the functions are divided between the two.

Next, referring to FIG. 4, the operation of this embodiment will be described in detail. First, the configuration of the shared file 31 existing on the secondary storage device 3, which is subject to simultaneous access control, will be described. The shared 7 isles 31 are hierarchically structured into blocks. Among these, a block, also called a page, is a unit in which input/output operations are performed, and is taken as an integral multiple of a physical input/output operation unit (depending on the device, for example, 256 bytes). On the other hand, a record is a logical access unit and is included in any block. Note that it is also possible to define records that span multiple blocks, but in such a case,
Each block based on the data is treated as a separate record. Here, every shared file 31 has a 7 stem to uniquely identify the file within the 7 stems. A file ID, which is another name for Air Ori, is attached. Each block also includes a block ID, which is a block identification name for uniquely identifying the block within the file to which it belongs. Further, each record is given a record ID, which is a record name used to uniquely identify the record based on the group to which it belongs. Below, what is block specification (file ID, block LD)?
Also, record specification refers to specifying a set of (file ID, block ID, record ID).

Next, the operation of simultaneous file access control will be explained along the flow of processing requests. First, a task that performs simultaneous file access makes the following request to the simultaneous file access control routine 24.

(1) When reading a record, specify the record and issue a GET request.

(2) When writing a record, specify the record and issue a PUT request.

(3) When a logical group of files has been updated, usually called a checkpoint section, Che, Kuboi/
Make a request.

The processing procedure in the file simultaneous access control routine 24 is as follows.

(1) GET Process ■ A lock request is made to the record sharing exclusive control means 11 in shared mode, specifying a record. At this time, if necessary, the record sharing exclusive control means 11 makes a wait.

(2) A buffer reservation request is made to the shared buffer management means 21, specifying the block containing the requested record. At this time, if necessary, the corresponding block is read into the shared buffer area 22 by the shared buffer management means 21.

(2) The requested record is transferred from the buffer obtained by the shared buffer management means 21 to the record storage area of the requesting task.

(2) A buffer release request is made to the shared buffer management means 21 in reference mode, specifying a processed block.

(2) PUT Process ■ A write request is made to the record sharing exclusive control means 11 in exclusive mode, specifying a record. At this time, if necessary, the record sharing exclusive control means 11 makes a wait.

(2) A buffer reservation request is made to the shared buffer management means 21, specifying the block containing the requested record. At this time, the corresponding block is read into the shared buffer area 22 by the shared buffer management means 21 if necessary.

(2) The requested record is transferred from the record storage area of the requesting task to the buffer obtained by the shared buffer management means 21.

(2) A buffer release request is made to the shared buffer management means 21 in update mode, specifying a processed block.

(3) Check and point processing ■ A buffer invalidation request is made to the shared buffer management means 21 for all buffers referenced by the requesting task.

(2) A request is made to the record sharing exclusive control means 11 to unlock all records locked by the requesting task.

In this embodiment, while a certain block is being read into the shared buffer area 22 of one computer 2, any record contained in that block is read.
The same block is protected from being read into the shared buffer area 22 of the other computer 2 under the control of the record sharing exclusive control means 11.

Although efficient allocation control of the shared buffer area 22 is also important as a function of the shared buffer management means 21, only the functions related to simultaneous file access control will be described here.

First, with reference to FIG. 5, a shared buffer management table 2 is a management table used by the shared buffer management means 21.
3 will be explained.

Each entry in the management table is created for each corresponding block, and includes block information consisting of a file ID and block ID, buffer area information consisting of a pointer to an area, an update flag, an in-use task list, and a reference task list. I'm here.

Next, the processing procedure for each request will be explained.

(1) Check whether the requested block already exists on the shared buffer management table 23. If it exists, it is referred to by a task that has already read the content.

If it does not exist, a free buffer area is allocated, the requested block is read, and the requesting task references it. In either case, the requesting task is registered in the buffer management entry's in-use task list and reference task list.

If the necessary free buffer area is not available, the buffer corresponding to the management entry whose busy task list is empty is released and allocated to the requested block. At this time, if the buffer has an update flag set, writing to the corresponding block is performed before allocation to the next block. Are there any buffers with no tasks in use?
Alternatively, if there is insufficient space in these buffers, a wait is performed until space becomes available.

(2) The buffer release processing request task is deleted from the in-use task list of the buffer management entry corresponding to the requested block.

If the request is in update mode, the update flag of the buffer management entry is set.

(3) The buffer invalidation processing request task is deleted from the reference task list of all buffer management entries. Here, for the buffer management entry whose reference task list is empty (at this time, there should be no task in use), if the update flag is set, after writing to the corresponding block. , en) IJ is deleted.

This buffer invalidation process is necessary in order to avoid holding in the shared buffer area 22 a block that is not locked for all the records it contains.

Finally, the operation of the record sharing exclusive control means 11 will be explained.

First, the shared exclusive control table 12, which is a control table used by the record shared exclusive control means 11 shown in FIG. 4, will be explained with reference to FIG. Resources to be controlled are managed in two levels: blocks and records. The first layer block control entry contains block information consisting of a file ID and block ID, a total usage width of 3E ID, a block waiting task queue, and a record control entry list corresponding to the records belonging to the block. Contains. The record control entry in the second hierarchy includes record information consisting of record 11), a list of currently waiting tasks, and a task key for waiting for records. Here, in the locked task list, record waiting task queue, and block waiting task queue, for each task in each state, computer II) of computer 2 to which the task belongs, task If), shared mode, exclusive mode The lock request type, such as the type of lock request, is registered. Here, computer II) is an identification name for uniquely identifying each computer 2 in the system, and task ID is an identification name for uniquely identifying each task within the corresponding computer 2.

Next, the processing procedure of the record sharing exclusive control means 11 will be explained with reference to a flowchart.

First, the procedure for processing an original request will be explained in detail below with reference to FIG.

The block has not been accessed by any task on Fine Needle 3Jfj&2 (corresponding to 71, 72), and there is no access conflict on a record basis with other tasks on the same computer 2. (corresponding to 73 and 74) are checked, and if a2 is allowed, the requesting task is registered in the pending task list of the corresponding record control entry. If the lock cannot be granted, the requesting task is registered at the end of the record-waiting task or block-waiting task, and is kept waiting until the lock becomes available.

Note that the double lock processing is the same as the conventional processing, so the processing flow is omitted. The existing lock type indicates shared, the lock type of the request indicates exclusive, and
If there are other tasks with shared lock A, it is necessary to wait until there are no more tasks with shared lock. Lock separation after double lock processing is shared if both existing and requested locks are shared, and exclusive in other combinations.

Next, regarding the unlock request processing methods 1i and 8,
This will be explained in detail with reference to the drawings.

First, the unlock request task is deleted from the locking task list of the corresponding record control entry, and if there are no more locking tasks, lock wait release processing is performed. In the record waiting release process 81,
From the head of the record waiting task queue of the record control entry to Iil, and in the block waiting release process 82, the same process as the locking process is performed again from the head of the block waiting task queue of the block control entry.

In the above description, only the shared exclusive control in units of records, which is a feature of the present invention, has been explained, but it is also possible to incorporate the conventional shared exclusive control in units of blocks into the shared exclusive control device l at the same time. In addition, as has been done in the past, before waiting for a lock request in the shared exclusive control device 1, a check is made to see if a dead-drop condition will occur if the waiting occurs, and if this occurs, the waiting is not performed. It is also possible to incorporate a de-lock avoidance mechanism that rejects requests in return.

〔Effect of the invention〕

As explained above, in the present invention, when controlling concurrent file access in a loosely coupled multiprocessor system, the hierarchy of blocks and records is taken into consideration, and if there is an access conflict between tasks running on the same computer, the sharing is done on a record-by-record basis. Exclusive control is performed. This reduces the frequency of waits due to access conflicts between tasks that access files simultaneously, and improves system processing power, compared to the conventional case where shared exclusive control is performed on a block-by-block basis. effective.

[Brief explanation of drawings]

Fig. 1 is a diagram showing an embodiment of the present invention, Fig. 2 is a diagram explaining simultaneous record-based file access control in a Kungle processor system, and Fig. 3 is a diagram showing simultaneous record-based file access without a shared buffer. 4 is a diagram for explaining the simultaneous file access control process in one embodiment. FIG. 5 is a diagram showing the shared buffer management table 23. FIG. 6 is a diagram showing the shared exclusive control table 12. FIG. 7 is a diagram showing lock processing, and FIG. 8 is a diagram showing unlock processing.・・・・・・Record sharing exclusive control means,
12... Shared exclusive control table, 2...
・Calculator, 21... Shared buffer management means, 2
2...Shared buffer area, 23...Shared buffer management table, 24...File simultaneous access control routine, 3...β Agent Patent attorney Hara Uchi (5, “Jin ＼−)′ ′yFJ1 Illustrated Answer Z Diagram Task 7/hizo 7a Task T?'s 8.177 Kikuda Rearrangement Z Diagram Foil δ Diagram

Claims (1)

[Claims] In a method for controlling simultaneous access to files on a secondary storage device shared by a plurality of computers constituting a loosely coupled multiprocessor system, a physical input/output unit is used as a control unit for the simultaneous access control. In a case where there is a block which is a control unit that is an integral multiple of a block, and a record which is a control unit that is smaller than a block, if there is an access conflict between tasks running on the same computer in response to a file access request in record units, Performs shared exclusive control on the requested record basis,
In addition, if there is an access conflict between tasks running on different computers, there is also a record sharing exclusive control means that performs shared exclusive control in units of blocks containing requested records, and a shared buffer area that exists for each individual computer. In the case of a request to allocate a buffer corresponding to a block, the shared buffer management means controls the allocation so that the same area is shared among multiple tasks requesting the buffer, and accesses the shared file in units of records. In,
File simultaneous access control in a loosely coupled multiprocessor system, characterized in that shared exclusive control is performed by the record shared exclusive control means, and physical input/output operations are performed via a buffer area allocated by the shared buffer management means. method.