JPH03230240A - Efficiency enhancing system for disk cache - Google Patents

Abstract

PURPOSE:To realize a disk cache with high efficiency by the small number of cache buffers by also using even unused pages as the disk cache. CONSTITUTION:An unused page queue 11 is provided to exert queue control onto the unused pages, and an unused page registering means 12 is provided to register the page storing a certain disk block to the unused page queue together with information to specify the page block. Then, an unused page retrieving means 13 is provided to retrieve the page storing the data of the designated disk block from the unused page queue, and a disk cache queue 14 is provided to exert the queue control upon the disk cache to store the data of the disk block. Further, a disk cache control means 15 to control the disk cache queue, and a disk 16 are provided. Therefore, the unused pages can be used as the disk cache. Thus, the disk cache with high efficiency is realized by the small number of cache buffers.

Description

DETAILED DESCRIPTION OF THE INVENTION [Field of Industrial Application] The present invention relates to a method for increasing the efficiency of a disk cache, and particularly to a method for increasing the efficiency of a disk cache in a computer system having a disk cache buffer in memory.

[Conventional technology]

Conventionally, disk cache has performed caching within the range of the number of cache buffers determined at the time of system generation.

[Problem to be solved by the invention]

The conventional method described above has the drawback that if too many cache buffers are prepared, the amount of memory that can be used decreases, and if the number of cache buffers is too small, the effectiveness of the cache is reduced.

[Means to solve the problem]

The disk cache efficiency method of the present invention is based on the size of a page, which is a management unit of a secondary storage device (hereinafter referred to as ``memory''), and the size of a secondary storage device (hereinafter referred to as ``memory'') set on the memory.
A disk cache queue whose unit size matches the unit size of the input/output cache buffer (hereinafter referred to as "disk cache") of the "disk"), a disk cache control means for registering, deleting, and searching the disk cache, and unused an unused page queue for queuing and managing pages; and an unused page registration means for registering pages used for disk cache in the unused page queue along with information identifying stored disk blocks. , an unused page search means for searching an unused page queue for a page in which a designated disk block is stored.

〔Example〕

Next, the present invention will be explained in detail with reference to the drawings.

FIG. 1 is a block diagram showing the configuration of an embodiment of the present invention, FIGS. 2(a) and (b) are block diagrams showing the configuration of an unused page queue and a disk cache queue in this embodiment, and FIG. The figure is a diagram showing the control flow of the unused page queue and disk cache queue in this embodiment.

In FIG. 1, 11 is an unused page queue for queue management of unused pages, and 12 is a page storing data of a certain disk block, which is registered in the unused page queue along with information identifying that disk block. unused page registration means; 13, unused page search means for searching an unused page queue for a page in which data of a specified disk block is stored; and 14, for queue management of a disk cache that stores data of a disk block. 15 represents a disk cache control means for managing the disk cache queue, and 16 represents a disk.

In Fig. 2(a) and (1), Fig. 2(a) shows an unused page queue, and 21 is the head pointer of the unused page queue, and 21a to 212 are connected to the unused page queue. This is an unused page management table. FIG. 2(b) shows a disk cache queue, and 22 (head pointers 22a to 22z of the disk cache queue) are disk cache management tables connected to the disk cache queue.

The management of unused page queues and disk cache queues will be explained in several cases with reference to FIGS. 1 to 3.

"Case 1" When a request is made to the disk cache control means 15 for a certain disk block, the disk cache control means 15 searches whether the requested disk block exists in the disk cache queue 14 and determines whether the requested disk block exists or not. If disk cache exists, remove Step 3
1) Remove the corresponding disk pro... from the disk cache queue.Step 32) Remove the corresponding disk pro...
The disk cache is returned together with the information indicating that it has been checked (step 39).

"Case 2" In step 31 of case 1, the corresponding disk block does not exist in the disk cache queue 14 (step 31), and the disk block requested by the unused page search means 13 is stored. Search whether the page exists in the unused page queue 11 (step 33), and if the page exists, remove the disk cache 22z at the end of the disk cache queue 14 from the queue (step 35),
The page used as a disk cache is removed from the disk cache, and the unused page registration means 12 connects it to the head 21a of the unused page queue (step 36).

If the page is invalid in step 35, the page used as the disk cache is removed from the disk cache, and the unused page registration means 12 connects it to the end 21z of the unused page queue (step 37). In either case, the page obtained in step 133 is removed from the unused page queue, connected to the disk cache obtained in step 34 (step 38), and the disk cache is returned together with the information that the corresponding disk block exists (
Step 39).

"Case 3J In step 33 of case 2, the corresponding page does not exist in the unused page queue 11, and if the disk cache 22z at the end of the disk cache queue 14 is removed from the queue (step 40), the removed disk character Check whether the queue is valid, and if so, remove the page (step 41), which is used as a disk cache, and remove the page used as a disk cache from the disk cache, and connect it to the head 21a of the unused page queue by the unused page registration means 12 (step 42). , removes the last page 21z from the unused page queue and stores it in the disk cache obtained in step 40 (step 43).
Regardless of whether it was valid or invalid in step 1, the information in step 4 that there is no corresponding disk block is displayed.
Return the disk cache obtained in step 0 (step 44)
).

〔Effect of the invention〕

As described in detail above, the present invention has the effect of realizing a highly efficient disk cache with a small number of cache buffers by using unused pages as a disk cache.

[Brief explanation of drawings]

FIG. 1 is a block diagram showing the configuration of an embodiment of the present invention, FIGS. 2(a) and (b) are block diagrams showing the configurations of an unused page queue and a disk cache queue, respectively, in the embodiment, and FIG. is a flowchart showing the control flow of an unused page queue and a disk cache queue in this embodiment. 11... Unused page queue, 12... Unused page registration means, 13... Unused page search means, 14.
Disk cache queue, 15...Disk cache queue
Soshu control means, 16...disk, 21...head pointer of unused page queue, 21a to 21Z...
Unused page management table, 22...Disk cache
Soshu cue head pointer, 22 a to 22 z...
- Disk cache management table.

Claims (1)

[Claims] The size of the page, which is the management unit of the primary storage device (hereinafter referred to as ``memory''), and the input/output cache buffer (hereinafter referred to as ``disk cache'') of the secondary storage device (hereinafter referred to as ``disk'') set on the memory. A disk cache queue whose unit size matches, a disk cache control means for registering, deleting, and searching disk caches, an unused page queue for queuing and managing unused pages, and a disk cache queue that matches the unit size. An unused page registration means that registers pages used for caching in an unused page queue along with information that identifies the disk block in which they are stored, and searches the unused page queue for a page in which a specified disk block is stored. A method for improving the efficiency of a disk cache, comprising: means for searching for unused pages.