JP2000353104A - Data backup system and its method - Google Patents

Abstract

PROBLEM TO BE SOLVED: To make it possible to quickly restart backup processing from a computer generating no fault even when one of computers executing backup processing runs down while using a small computer system interface(SCSI) bus. SOLUTION: In a computer system for allowing plural computers 1 to 4 to use a tape device 5 shared by exclusive control through an SCSI bus interface as a backup file device, computers 2, 3, 4 are concerned with the execution of backup. The computer 1 is provided with a backup monitoring part 6 for monitoring time required for the backup processing of the computer executing backup operation, and at the time of detecting a state that the execution of backup processing is not ended even after the lapse of prescribed time, selects the computer not generating a fault out of the computers 3, 4, allows the selected computer to reset the SCSI bus and restart the backup processing after resetting the SCSI bus.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a computer system backup method and method, and more particularly, to a SCSI system.
The present invention relates to a backup method for a computer system that uses a tape device shared by exclusive control using a (Small Computer System Interface) as a backup device when a failure occurs.

[0002]

2. Description of the Related Art Conventionally, as an interface on hardware for connecting peripheral devices of a small computer,
An interface that is commonly called a “scuzzy”, that is, SCSI is widely used.

[0003] However, a plurality of computers use the SCS.
In a conventional computer system using a tape device shared by I and exclusive control as a backup file device, one of the computers being backed up fails and the computer goes down while using the SCSI bus. In the backup processing in such a case, a trouble such as abnormal termination of a hang-up of the SCSI bus has occurred.

When the prior art related to the field of the present invention is retrospectively examined from past patent applications, first, Japanese Patent Application Laid-Open No. 4-27
No. 6843 discloses that in a SCSI system, SC
If one of the SCSI devices connected to the SI bus has the busy signal turned on and the time without change of the SCSI bus request signal exceeds a predetermined time, the SCS
A SCSI bus monitoring device that turns on an I bus reset signal to release a locked state is disclosed.

Japanese Patent Laid-Open Publication No. Hei 7-160618 discloses a method for recovering from a situation in which a shared SCSI bus hangs up.
There is disclosed a data processing device that shares information on the use state of a CSI device and, when there is illegal occupation, performs a bus reset to release the bus from illegal occupation.

[0006]

However, in the prior art, as described above, a plurality of computers use SCSI,
In a computer system in which a tape device shared by exclusive control is used as a file device for backup, if one of the computers being backed up fails and goes down while using the SCSI bus, another computer is used. Even when used, there was a problem that the backup process could not be performed due to the hang-up of the SCSI bus.

In addition, when another computer restarts during backup, the SCSI bus is reset and the backup ends abnormally.

The technique disclosed in Japanese Patent Application Laid-Open No. 4-276843 requires a new SCSI bus monitoring device to be connected to the SCSI bus, separately from a group of computers for practical use. . In addition, it is not a technology related to sharing a file device for backup.

The technique disclosed in Japanese Patent Application Laid-Open No. Hei 7-160618 discloses a technique for accessing a device by using a SCSI.
If the bus is busy, then determine if it is illegal occupancy,
Since the bus reset is performed, there is a problem that it is not possible to detect that the SCSI bus has hung up until the bus is accessed. Also, the processing that was being performed when the hangup occurred was not considered. Further, it is not a technology relating to sharing of a backup file device.

The present invention has been made in view of the above-described problems in the conventional computer system backup system. One of the computers being backed up has a failure, and the SCSI bus is used. It is an object of the present invention to provide a computer system backup method and a backup method capable of promptly re-executing a backup from a computer in which no failure has occurred even if a failure occurs.

[0011]

SUMMARY OF THE INVENTION A data backup system according to the present invention controls a bus, a file device connected to the bus for backing up data, and controlling the bus. First and second writing means for writing data to the file device, backup execution means for giving data to the first writing means and instructing the writing of the data, and hang of the first writing means Monitoring means for monitoring an up time by time measurement, and issuing a rewrite instruction to the backup execution means when a hang-up is detected; and a means for resetting the bus when a hang-up is detected. The backup executing means gives the data to the second writing means when receiving the rewriting instruction. Characterized by instructing the writing of the data Te.

Further, the data backup method according to the present invention provides a bus, a file device connected to the bus for backing up data, controlling the bus, and controlling the file device via the bus. Writing means for writing data, backup executing means for giving data to the writing means and instructing the writing of the data, and detecting means for detecting reset of the bus, wherein the reset of the bus is detected. When writing, the writing means rewinds the file device, and the backup executing means gives the data to the writing means and instructs the writing of the data.

Further, in the data backup system according to the present invention, in the above data backup system, the bus is a SCSI (Small Computer System Interface) bus.

Further, in the data backup system according to the present invention, in the above-mentioned backup system, the file device is a tape device.

The data backup method according to the present invention comprises:
A step in which the backup executing means gives data to the first writing means and instructs writing of the data; a step in which the first writing means writes the data to a file device via a bus; Monitoring the hang-up of the means by time measurement, resetting the bus when the hang-up is detected, and when the hang-up is detected, the backup execution means is connected to the second writing means. The method further comprises the steps of: giving the data and instructing the writing of the data; and writing the data to a file device via a bus by the second writing means.

Also, in the data backup method according to the present invention, the backup executing means gives the data to the writing means and instructs the writing of the data, and the writing means writes the data to the file device via the bus. Detecting the reset of the bus, when the reset of the bus is detected, the step of rewriting the file device by the writing means, and the backup execution means giving the data to the writing means again. And writing the data to the file device via the bus again by the writing means.

Further, in the data backup method according to the present invention, in the above data backup method, the bus is a SCSI bus.

Further, in the data backup method according to the present invention, in the above data backup method, the file device is a tape device.

[0019]

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

FIG. 1 is a block diagram showing an overall configuration and an overall operation of a backup system of a computer system according to an embodiment of the present invention.

The backup method of the computer system according to the present embodiment includes a computer 1 configured by computers operated under program control, a computer 2 connected to the computer 1, and a computer 2 connected to the computer 1 and the computer 2. Computers 3, 4
And a tape device 5 connected to the computers 3 and 4 via the SCSI bus 12 and shared by the computers 3 and 4 by exclusive control.
And

The computer 1 includes a backup monitor 6.

The computer 2 includes a backup execution unit 7.

The computer 3 includes a tape writing unit 8 and a failure detection / recovery unit 10, and the computer 4 includes a tape writing unit 9
And a failure detection and recovery unit 11.

The operation of each of the above components will be described below.

The backup monitor 6 receives the backup start time from the tape writer 8 or the tape writer 9 and monitors the backup time. If the above monitoring causes a state in which the backup does not end even after a certain period of time (in the meantime, the occurrence of such a state can be detected by keeping the SCSI bus busy), the failure detection recovery The SCSI bus 12 is reset using the unit 10 or the failure detection / restoration unit 11.

The backup execution unit 7 transfers the backup data to the tape writing unit 8 or the tape writing unit 9.

The tape writing units 8 and 9 receive the backup data from the backup execution unit 7 and write the received data to the tape device 5.

When detecting that the SCSI bus 12 has been reset, the failure detection / restoration units 10 and 11 notify the backup execution unit 7 of the detection result. Further, upon receiving the notification from the backup monitoring unit 6, the SCSI bus 12 is reset.

Hereinafter, the overall operation of the backup system of the computer system according to the present embodiment will be described with reference to FIG.

First, in step A1, the backup data is transferred from the backup execution unit 7 to the tape writing unit 8.

Next, in step A2, the tape writing unit 8 that has received the backup data
Start backup at At the same time, step A
In 3, the tape writing unit 8 notifies the backup monitoring unit 6 that the backup has started. The notification is accompanied by the identifier of the computer 3. The backup monitoring unit 6 monitors the time, and if the backup is not completed for a certain period of time,
In step A4, a request is made to the failure detection / restoration unit 11 to perform a SCSI bus reset. Since the reset request destination computer is a computer having an identifier different from the identifier accompanying the request in step A3, the reset request destination is the computer 4.

In step A5, the failure detection / restoration unit 11 having received the notification executes a SCSI bus reset for the tape device. By performing the SCSI bus reset, the access right to the tape device 5 by the tape writing means 8 is released.

Thereafter, in step A6, the backup monitoring unit 6 notifies the backup execution unit 7 of a backup re-execution instruction. This notification also includes step A3
Accompanied by the identifier associated with the request in.

In step A7, the backup executing unit 7 which has received the backup re-execution instruction transfers the backup data to the tape writing unit 9 of the computer 4 where no failure has occurred. Since the transfer destination computer is a computer having an identifier different from the identifier accompanying the request in step A3, the reset request destination is the computer 4.

In step A8, the tape writing unit 9 that has received the backup data starts backup to the tape device 5. At the same time, in step A9, the tape writing unit 9 notifies the backup monitoring unit 6 that the backup has started. When the above backup is completed normally, in steps A10 and A11,
The tape writing unit 9 notifies the backup monitoring unit 6 and the backup execution unit 7 that the backup has been completed normally.

The tape writing section 8 and the tape writing section 9 have the same SCSI ID (usually 0), but are controlled by the backup monitoring section 6 and the backup execution section 7 as described above. There is no contention on the SCSI bus.

FIG. 2 is a block diagram showing an operation when a SCSI bus reset is detected during backup in the computer system backup method according to the embodiment of the present invention.

The operation of the computer system according to the present embodiment when the SCSI bus reset is detected during the backup will be described below with reference to FIG.

Steps A1 to A3 are steps for starting backup, and this part is the same as the operation up to step A3 shown in FIG. Hereinafter, step A
The operation when the SCSI bus reset is detected by the failure detection / restoration unit 10 in step B4 during backup at step 3 will be described.

In step B5, the failure detection and recovery unit 10
Informs the tape writing unit 8 that the SCSI bus reset has been detected.

In step B6, the tape writing unit 8
Issues a rewind command to the tape device 5, and
In step B7, the backup monitoring unit 6 is notified that the backup has been stopped. The notification that the backup has been stopped is accompanied by the identifier of the computer 3.

In step B8, the backup monitoring unit 6
Notifies the backup execution unit 7 of a backup restart instruction. The notification instruction is accompanied by the identifier of the computer 3 accompanying the notification in step B7.

Thereafter, in steps B9 and B10, the backup executing section 7 resumes the backup using the tape device 5 as a backup file via the tape writing section 8.

In the example of FIG. 1, the failure detection / restoration unit 10 detects the SCSI bus reset by the failure detection / restoration unit 11 in step A5. However, since the tape writing unit 8 is hung up, It does not issue a rewind instruction or notify the backup monitoring unit 6 of the suspension of the backup.

In the example of FIG. 1, the failure detection / restoration unit 11 does not detect the SCSI bus reset by the failure detection / restoration unit 11 in step A5. Even if it is detected, there is no problem since the rewind instruction is performed by the tape writing unit 9 that subsequently writes data, and the backup stop notification to the backup 6 is accompanied by the identifier of the computer 4. There is no contradiction between the identifier associated with the notification of cancellation and the identifier associated with the notification in step A3, and the backup monitoring unit 6 can normally determine the computer to be used for resuming backup.

In the above embodiment, the computers 3 and 4 that write data to the tape device 5 are configured to transfer the backup data from the backup execution unit 7 of the computer 2. In general, since it is a backup file device shared by exclusive control,
It is possible to support backup of a multi-computer system of another arbitrary configuration.

In the above embodiment, a tape device is used as a backup file device. However, any file device that can be exclusively controlled can be used as a backup file device. Generally, any other exclusive controllable file device including a disk device can be used as a backup file device.

[0049]

According to the present invention described above, in a computer system in which a plurality of computers use SCSI and a tape device shared by exclusive control is a file device for backup, one of the computers being backed up is used. In the event that a failure occurs and the system goes down while using the SCSI bus, the above-described SCSI bus reset is executed from another computer. Can be implemented.

Thus, a tape device shared by exclusive control can be used as a backup file device without any trouble.

Also, during the backup, S
Even when the CSI bus is reset, the backup can be re-executed by once rewinding the tape device which is the backup file device.

[Brief description of the drawings]

FIG. 1 is a block diagram showing an overall configuration of a backup system of a computer system according to an embodiment of the present invention and an operation when a hang-up occurs.

FIG. 2 is a block diagram showing an operation when a SCSI bus reset is detected during backup in the computer system backup method according to the embodiment of the present invention.

[Explanation of symbols]

 DESCRIPTION OF SYMBOLS 1 Central processing unit 2 Processor 3 and 4 Data processing unit 5 Tape device 6 Backup monitoring unit 7 Backup execution unit 8, 9 Tape writing unit 10, 11 Failure detection and recovery unit

Claims (8)

[Claims] 1. A bus, a file device connected to the bus for backing up data, and a first and a second device for controlling the bus and writing data to the file device via the bus. A second writing unit, a backup execution unit that gives data to the first writing unit and instructs the first writing unit to write the data, and monitors a hang-up of the first writing unit by measuring a time. Monitoring means for issuing a rewrite instruction to the backup execution means when detected, and means for resetting the bus when a hang-up is detected, the backup execution means comprising: Receiving the data, giving the data to the second writing means, and instructing the writing of the data. Data backup method. 2. A bus, a file device connected to the bus for backing up data, and writing means for controlling the bus and writing data to the file device via the bus. Backup execution means for giving data to the writing means and instructing the writing of the data; anddetection means for detecting a reset of the bus.When the reset of the bus is detected, the writing means A data backup method wherein the file device is rewinded, and the backup executing means gives the data to the writing means and instructs the writing means to write the data. 3. The data backup method according to claim 1, wherein the bus is a SCSI (Small Compute
(r System Interface) A data backup method characterized by a bus. 4. The data backup method according to claim 1, wherein the file device is a tape device. 5. A step in which the backup executing means gives data to the first writing means and instructs writing of the data, and the first writing means writes the data to a file device via a bus. Monitoring the hang-up of the first writing unit by time measurement, resetting the bus when the hang-up is detected, and executing the backup execution unit when the hang-up is detected. 2. A data backup comprising: a step of giving the data to a second writing unit to instruct the writing of the data; and a step of the second writing unit writing the data to a file device via a bus. Method. 6. A step in which the backup executing means gives data to the writing means to instruct the writing of the data, the writing means writes the data to a file device via a bus, and detects a reset of the bus. And when the bus reset is detected, the writing means rewinds the file device, and the backup executing means gives the data to the writing means again to instruct the writing means to write the data. A data backup method, comprising: writing the data to a file device via a bus again by the writing means. 7. The data backup method according to claim 5, wherein said bus is a SCSI bus. 8. The data backup method according to claim 5, wherein the file device is a tape device.