JPH0237418A - Data matchability control system for storage device - Google Patents

Abstract

PURPOSE:To reduce the control table volume and the control program volume and to accept an N-fold write request during troubling by dividing contents stored in a storage device by an arbitrary unit and providing a difference bit map table indicating whether postwrite data reflects on the storage device or not with respect to each unit. CONSTITUTION:When both of N-fold write and postwrite are designated, data is stored in a cache 4, and respective bits of a difference bit map table 6 corresponding to write destinations of N-fold write storage devices 7 and 8 are set, and respective corresponding bits are cleared when data stored in the cache 4 is written in storage devices 7 and 8. N-fold write and postwrite functions are managed by the difference bit map, and unitary management is possible. Since difference information for all data in N-fold write storage devices 7 and 8 is managed by the difference bit map table 6 in the N-fold write mode, a two-fold write request can be accepted against the trouble, by which data cannot be written in a part of N-fold write storage devices 7 and 8, by reflection on the difference bit map table 6.

Description

DETAILED DESCRIPTION OF THE INVENTION [Field of Industrial Application] The present invention relates to a storage device control system in a buffered storage control device having N-overwrite and post-write functions.

[Conventional technology]

Conventionally, related methods for controlling the N-overwrite function in storage control devices include, for example, Japanese Patent Application Laid-Open No. 61-208119.
No. 61-249132, No. 61-208119, etc.

Further, related methods for controlling the afterwrite function include, for example, Japanese Patent Laid-Open No. 61-240320.

(Problems to be Solved by the Invention) The above-mentioned conventional technology does not take into consideration the unified management of N-overwrite and post-write functions, and is realized using separate management systems, resulting in problems with the amount of control tables and control programs. Also, during N-overwrite mode, due to failure or maintenance, N
No consideration has been given to handling N overwrite requests when data from the host interface can only be written to one side of the overwrite secondary storage device, and N during the failure
There was a problem with overwriting requests being rejected.

The purpose of the present invention is to centrally manage N-overwrite and post-write functions,
The object of the present invention is to reduce the amount of control tables and control programs, and to make it possible to accept N overwrite requests during the failure.

[Means to solve the problem]

The above purpose is to provide a storage controller with a cache with a differential bitmap table that stores information on whether data received from the host interface is reflected for each unit of memory in the storage device and an N-overwrite. This is achieved by providing a pair information table that stores the identification numbers of paired storage devices.

[Effect]

When there is a request for N-multiple writing, the write data is written to one of the N-multiple writing storage devices and written to the cache at the same time. At this time, a bit in the differential bitmap table corresponding to the write destination of the next unwritten storage device is set, and the bit is cleared when the data stored in the cache is written to the storage device. do.

Also, when there is a write-after request, the data is stored in the cache for write-later, the bit of the differential bitmap table corresponding to the write destination in the storage device is set, and the data stored in the cache is written to the storage device. When the bit is cleared, the corresponding bit is cleared.

When both N-overwrite and post-write are specified, the data is stored in the cache, each bit of the differential bitmap table corresponding to each write destination of the N-overwrite storage device is set, and the data is stored in the cache. Clear each corresponding bit when data is written to the storage device.

The existence of N-overwrite or post-write data stored in the cache is determined by referring to the differential bitmap.

As a result, the N-overwrite and post-write functions can be managed using differential bitmaps, so that they can be combined.

When in N overwrite mode, N is determined by the differential bitmap table.
Difference information is managed for all data in the overwrite storage device, so in the event of a failure where part of the N overwrite storage device cannot write data, the double write request will be reflected in the differential bitmap table. By doing so, it can be accepted.

〔Example〕

Embodiments of the present invention will be described below with reference to FIG.

First, the present invention will be explained in detail. 1 is a magnetic disk control device, 2 and 3 are directors that control the magnetic disk device, cache, differential bitmap table, and counter information table, 4 is a cache that is also used as a data buffer to store double-written data, and 5 is a nonvolatile 6 is a differential bitmap table that stores information on whether or not the data received from the host interface is reflected for each track in the magnetic disk device; 7 and 8 are a dual-write pair; The magnetic disk device 9 is a magnetic disk device.

101 is a host interface line for transferring commands, status, data, etc. to the host, and 102 and 103 are magnetic disk devices! 7, 8.9 and the magnetic disk control device 1, and is a device interface line for transferring commands, status, data, etc.

Next, each operation will be explained using a case of double writing as an example.

In the dual write mode of operation (FIG. 2), the write data from the host interface to the host interface line 101
When the data is received by the director 2 via the device interface line 102, the data is written to the magnetic disk device 7 via the device interface line 102, and at the same time, the data is stored in the cache.
The bit in the differential bitmap table 6 corresponding to the write destination track of the partner magnetic disk device 8 that is a pair of the dual-write magnetic disk device is set.

Upon a certain activation, the director 3 refers to the differential bitmap table 6 and writes the magnetic disk device 8 . Write the data in the cache that has not yet been reflected in the data from the host interface for the track in the track 9 to the magnetic disk device FI8.9 via the device interface line 103;
Then, the data is reflected on the magnetic disk device 8.

Operation 2 In the write-after mode (FIG. 3), when write data is received by the director 2 from the host interface via the host interface line 101, the data is stored in the cache 4.

Upon a certain activation, the director 2 refers to the differential bitmap table 6 and writes the data in the cache 4, which is data received from the host interface for the tracks in the magnetic disk devices 7 and 9 and has not yet been reflected, via the device interface line 102 to the magnetic disk. The data is written to the disk drives W7 and W7, and the data is reflected in the magnetic storage device 9.

Operation 3 In the dual write and post write mode (FIG. 4), when the director 2 receives write data from the host interface via the host interface 1lo1, the data is stored in the cache and the dual write magnetic disk device 7,
The bits of the differential bitmap table 6 corresponding to the write destination tracks of 8 are respectively set.

At a certain startup, the directors 2 and 3 refer to the differential bitmap table, respectively, and transfer the unreflected data in the cache received from the host interface for the tracks in the magnetic disk devices 7 and 8.9 to the device interface line 102. .103 to the magnetic disk drives 7, 8 and 9, and the data is reflected in the magnetic disk drives 7, 8.

Operation 4 When a cache failure occurs in dual write mode (5th
When the director 2 receives write data from the host interface via the host interface line 101, it writes the data to the magnetic disk drive 7 and writes it to the write destination track of the magnetic disk drive 8 of the partner for double writing. Set the corresponding bit in the differential bitmap table 6. Writing of the data to the partner magnetic disk device 8 is not performed during a cache failure.

When the cache is restored, at a certain startup, the director 2 refers to the differential bitmap table 6 and installs the magnetic disk W.
Correct data is reflected in the magnetic disk unit W7, but the track data of the magnetic disk unit W7, which is not reflected in the magnetic disk unit 8 and is not stored in the cache, is transferred to the device interface line 1.
The data is stored in the cache from the magnetic disk device 7 via 02.

Thereafter, the director 3 writes the data to the magnetic disk unit Wt8 via the device interface line 103, as in the normal double write mode.

Also, from the viewpoint of reliability, the nonvolatile shared memory is duplicated.

According to this embodiment, it is possible to centrally manage the dual write and postwrite functions, it is possible to accept double write requests from the host interface during a cache failure, and it is possible to accept double write requests from the host interface during a cache failure, and it is possible to perform double write requests between dual write secondary storage devices during failure recovery. This has the effect of speeding up data matching and making it possible to continue double writing.

〔Effect of the invention〕

According to the present invention, consistency regarding N-overwrite and post-write functions can be centrally managed, which has the effect of reducing the amount of control programs. Further, even if the N-overwrite data cannot be reflected in some storage devices of the N-overwrite storage device due to a failure or maintenance, the N-overwrite request can be accepted, which has the effect of allowing business to continue. After recovery from a failure or after completion of maintenance, the difference bitmap is used to quickly match data between N-overwrite storage devices, and the N-overwrite mode can be continued.

[Brief explanation of the drawing]

FIG. 1 is a block diagram of the engineering/○ subsystem of the embodiment of the present invention, and FIGS. 2, 3, 4, and 5 explain the management of the differential bitmap table in the embodiment of the present invention. FIG. 1... Magnetic disk control device, 2, 3... Director, 4... Cache, 5... Non-volatile shared memory, 6... Differential bitmap table, 7, 8... Dual writing magnetic disk device, 9... magnetic disk device,
101...Host interface line, 102, 103
... Figure 3

Claims (1)

[Claims] 1. In a system comprising a storage device and a storage control device with a buffer (cache) having a post-write function to the storage device, the storage is divided into arbitrary units in the storage device,
1. A data consistency control method for a storage device, characterized in that a differential bitmap table is provided that indicates whether postwritten data is reflected in the storage device for each unit. 2. In the above system, in order to realize the N overwriting function (N is an integer of 2 or more), data between N overwriting storage devices is calculated from the differential bitmap table corresponding to N units of N overwriting storage devices. 2. The data consistency control method for a storage device according to claim 1, wherein the presence or absence of a difference is determined by calculation. 3. In the above system, during the N-overwrite mode, write requests to the storage device are accepted for each of the N-overwrite storage devices, and a differential bitmap corresponding to the write destination of the partner storage device in the N-overwrite is created. 3. The data integrity control method for a storage device according to claim 2, further comprising a mode in which a bit in the table is set and the data is not written to the partner storage device. 4. Regarding the above system, if it becomes impossible to write N multiple times due to a failure or maintenance, write data only to one storage device, do not write data to other storage devices, and operate in the mode set forth in claim 3 above. A claim characterized in that a mode is provided for rapidly matching data between N overwrite storage devices by referring to a differential bitmap table and copying only the data in which a difference exists after failure recovery or maintenance is completed. A data integrity control method for a storage device according to item 3. 5. The data consistency control method for a storage device according to claim 4, wherein in the above system, a mode is provided in which N-overwriting is continued at the time of failure recovery or maintenance completion. 6. The data integrity control device for a storage device according to claim 5, wherein in the system, the differential bitmap table is allocated to a non-volatile memory.