JPS6117014B2 - - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION The present invention provides a historical information storage control method, and particularly a system that stores the same historical information by forming a set of at least two external storage devices such as disk devices (hereinafter, a disk device will be explained as an example). In a history information storage system configured to store redundant data, the history information collection method using healthy disk devices is rearranged according to the manner in which failures occur in each disk device, in order to avoid system failures as much as possible. The present invention relates to a history information storage control method.

Conventionally, particularly in on-line processing systems, historical information about processing performed by the system is collected on a historical file, such as on a disk medium. In the collection, a plurality of disk devices are arranged in advance into groups of at least two, and each group is used in circulation. That is, as shown in FIG. 1A, the disk devices (V ₁ , X ₁ ) and (V ₂ ,
a first set with disk devices (V ₂ , X ₂ ) and (V ₃ , Y ₂ ); a _second set with disk devices (V 3 , Y ₂ );
A third set includes X ₃ ) and (V ₄ , Y ₃ ), and a fourth set includes disk devices (V ₄ , X ₄ ) and (V ₁ , Y ₄ ). In a state where the same history information is being collected redundantly using one set, when the disk device of the set becomes full, the second set is prepared for use next. In this _{case ,} as shown in FIG.
A configuration using a set of is adopted.

However, if the disk device indicated by the asterisk becomes a failure as shown in FIGS. 2A and 2B, it becomes impossible to use the above-mentioned predetermined set, leading to the system going down. In particular, there has recently been a tendency to adopt a method of fixedly allocating disk media to disk devices, making it difficult to maintain the above-mentioned system by replacing disk media. It becomes necessary.

In order to solve the above-mentioned problems, the present invention aims to automatically reconfigure the groups shown by diagonal lines in FIGS. 2A and 2B to prevent the system from going down. Furthermore, it is intended to provide flexibility so that the system can be operated even with a healthy disk device alone, as shown in FIG. 2C. Therefore, in the history information storage control method of the present invention, a plurality of external storage devices are connected to each of a plurality of input/output control devices, and the plurality of external storage devices are configured in advance into groups of at least two. and is configured to store history information about processing performed by the system in the one set, and the upper control device of the plurality of input/output control devices stores the history information. In a data processing system configured to specify a write destination of the history information at a time, the upper control device determines the mode of failure occurrence in each of the external storage devices and sets the usage mode of the external storage device. The file management section includes a failure mode determination section and a new set corresponding to the mode of failure occurrence in which the predetermined set cannot be used, and the file management section reconfigures a new set corresponding to the mode of failure occurrence in which the predetermined set cannot be used, and at least the same input/output control device. The present invention is characterized in that the external storage devices connected to the external storage devices can be used as one set, and includes a chain generation section that generates a chain between healthy external storage devices. This will be explained below with reference to the drawings.

Figure 3 shows the configuration of one embodiment of the present invention, and Figure 4 shows the configuration of the third embodiment.
The configuration of one embodiment of the file management section shown in the figure, FIG. 5A shows the state of the file management section that constitutes the aspect shown in FIG. 2A, and FIG. 5B shows the file management that constitutes the aspect shown in FIG. 2B. Indicates the condition of the section.

In FIG. 3, 1 is a host control device, 2-0,
2-1 is a disk control device, 3 to 10 are disk devices, 11 is a write area designation unit,
Reference numeral 12 denotes a data management unit that manages data transferred to and from disk devices, 13 a failure detection unit that detects individual failure states of each disk device 3 to 10, and 14 a file controller. A management unit that determines the failure status of each disk device and instructs usable disk devices; 1
Reference numeral 5 represents a higher-level control device control section.

Upon receiving a write command W from the host processing device,
The write area designation section 11 is connected to the file management section 14.
Specify the writing area based on information from. That is, specify which disk device the history information WD is to be written to. In response to this, the data management unit 12 controls writing to the disk devices 3 and 7 as a pair, as shown in FIG.
10 to 10 are checked for failure occurrence status. The failure detection unit 13 detects the occurrence of a failure in each disk device 3 to 10 through the data management unit 12.
When the file is detected based on information from the file management section 14, the file management section 14 is notified of this fact. Based on this, the file management section 14 prepares a system for using the disk device as shown in FIG. 1B, FIG. 2A, FIG. 2B, or FIG. 2C. That is, the next disk device to be used is prepared and placed on standby.

FIG. 4 shows an embodiment of the configuration of the file management section 14 shown in FIG. In the figure, reference numeral 16 is a disk device failure occurrence storage unit, 17 is a failure occurrence mode determination unit, 18 is a chain generation unit, 19 is a pair use chain, 20 is a pair use chain pointer, 2
1 is a pointer for a failed disk device chain, 22 is a pointer for a failed disk device chain, 23 is a free disk device chain, 24 is a pointer for a free disk device chain, 25 is a disk device single use chain, and 26 is a pointer for a disk device single use chain. It represents.

The illustrated disk device failure occurrence storage unit 16 stores information corresponding to the volume number (V) and file name (X) assigned to each disk device 3 to 10.
Stores the failure occurrence status of the corresponding disk device. In the illustrated case, the disk device 4 indicated by (V ₂ , X ₂ ) is shown as having a failure. The fault occurrence mode determination unit 17 knows that the pair relationship shown by the arrow “pair” in the storage unit 16 has been given in advance by the user, and based on this knowledge, it determines the occurrence of the problem in each disk device 3 to 10. Determine the manner in which the failure occurred.
The chain generation unit 18 also generates chains 19, 21, 23, and 25 based on the determination results.

In the illustrated _case , _the chain 19 corresponds to the failure occurrence mode _illustrated in FIG _. A third set of devices (V ₃ , X ₃ ) 5 and (V ₄ , Y ₃ ) 9, disk devices (V ₄ , X ₄ ) 6 and (V ₁ , Y ₄ ) 10
The fourth set with , and the pointer (P ₁ ) 2
The 0 indicates that the third set should be used next. Further, a disk device (V ₃ , X ₂ ) 4 is chained to the chain 21, and a disk device (V ₃ , Y ₂ ) 8 is chained to the chain 24. Further, the chain 25 is adapted to deal with a special abnormal state by chaining the disk devices (V ₃ , Y ₂ ) 8 which are in an empty state.

FIG. 5A shows the state of the chain generating section 18 corresponding to the failure occurrence mode as shown in FIG. 2A. In this _case , the failure occurrence mode determining unit ₁₇ shown in FIG.
Based on the failure of (V ₄ , Y _{3 ) 9 and (V 4 , X 4 ) 10 ,} disk devices ( _{V 3} , It is determined that there is a need to reconstruct the set such as . The chain generation unit 18 generates a set of disk devices (V ₁ , X ₁ ) 3 and ( _{V 2} , Y ₁ ) 7 as a chain 19, and a set of disk devices (V _{3 ,} Y ₄ ) Make a chain by pairing with 10.

FIG. 5B shows the state of the chain generating section 18 corresponding to the failure occurrence mode as shown in FIG. 2B. In this case, the determination is made in the same manner as above, and the chain 19 includes disk devices (V ₁ , X ₁ ) 3 and (V ₂ ,
pair with Y ₁ ) 7, and disk device (V ₃ , Y ₂ ) 8
and (V ₄ , Y ₃ )9 to form a chain.

In addition, in the case shown in FIG. 2A, creating a set with disk devices (V ₃ , X ₃ ) 5 and (V ₃ , Y ₂ ) 8 means that the disk devices have the same volume number. Therefore, it is not necessarily a desirable form.

Furthermore, in the case shown in FIG. 2C, one set of disk devices (V ₁ , X ₁ ) 3 and (V ₂ , Y ₁ ) 7 remains on the chain 19 shown in FIG. The disk devices (V ₁ , Y ₄ ) 10 made above constitute one chain as a whole. As shown in FIG. 2C, if a disk device is used in a standalone state, for example, if one of the disk devices currently being used in a set fails, the remaining healthy side It is also possible to continue operating the disk device alone and, when it becomes full, to take over the storage process to the next group or the next disk device.

As explained above, according to the present invention, compared to the conventional method of retaining the form of a predetermined set and then skipping that set and selecting the next set in the event of a failure, the set is used cyclically. This makes it possible to collect history information using disk devices that are as healthy as possible. Then, a desired chain is generated in advance depending on the manner in which a failure occurs, so that when it becomes necessary to switch disk devices, the switching process can be performed appropriately.

[Brief explanation of the drawing]

1 and 2 are explanatory diagrams explaining the prerequisite problems of the present invention, FIG. 3 is an embodiment of the configuration of the present invention, and FIG. 4 is an embodiment of the configuration of the file management section shown in FIG. 5A shows the state of the file management section configuring the embodiment shown in FIG. 2A, and FIG. 5B shows the state of the file management section configuring the embodiment shown in FIG. 2B. In the figure, 1 is a host control device, 2 is a disk control device, 3 to 10 are disk devices, 11 is a write area designation section, 12 is a data management section, 13 is a failure detection section, 14 is a file management section, and 17 is a The failure mode determination section 18 represents a chain generation section.

Claims (1)

[Claims] 1 Processing performed by a system in which a plurality of external storage devices are connected to a plurality of input/output control devices, and the plurality of external storage devices are configured in advance into groups of at least two. is configured to store history information about the above in the one set, and when the higher-level control device of the plurality of input/output control devices stores the history information, it specifies a write destination of the history information. In the data processing system configured, the upper control device includes a file management unit that determines the manner in which a failure occurs in each of the external storage devices and sets the manner in which the external storage device is used; The occurrence mode determining unit and the above predetermined set are reconfigured to correspond to the mode of failure occurrence in which the above predetermined set cannot be used, and at least one set of external storage devices connected to the above-mentioned same input/output control device is created. 1. A history information storage control method characterized by comprising a chain generation unit that generates a chain between healthy external storage devices.