JPS62115555A - Managing method for duplexing auxiliary storage device of computer system - Google Patents

Abstract

PURPOSE:To improve the reliability by providing a state detection means for a duplexed auxiliary storage device (bulk), an operation supervisory means and a duplexed management means to restart or interrupt the duplexed operation of the bulk automatically without intervention of the operator. CONSTITUTION:An auxiliary storage state detection section 4, an auxiliary storage device operation supervising section 5 and an auxiliary storage device duplexing management section 6 are provided in a CPU 1. The detection section 4 detects the ready and non-ready state of the duplexed auxiliary storage devices 2, 3 (bulk), the supervisory section 5 is started at generation time of input/output error and when the bulk state is detected by the detection section 4, and the management section 6 manages the bulk. That is, when one of the bulk is failed by the detection section 4, the access to the bulk is intermitted by the supervisory section 5 and the management section 6 and only the other sound bulk is used to keep the operation. When the fault is restored, the content of the other sound bulk is copied on the restored bulk, then the duplicated operation of the bulk is restarted thereby improving the reliability of the system.

Description

DETAILED DESCRIPTION OF THE INVENTION [Technical Field of the Invention] The present invention relates to a redundant auxiliary storage management method for a computer system.

[Technical background of invention 1! i] Normally, with regard to dual auxiliary storage devices (hereinafter referred to as auxiliary storage devices) that match data using dual writes, if one of the dual bulk devices breaks down or When disconnected from the system for maintenance/inspection, etc., and then reconnected to the system after the failure is recovered or maintenance/inspection is completed, the system must be stopped, or operated in an online state according to instructions from the operator. Failure, recovery, maintenance and
The method used was to copy the inspected bulk to match the contents and restart duplex operation again.

[Problems with the background art] According to the conventional method described above, operator intervention is required, and if the direction of copying is incorrect due to an erroneous operation, the contents of the bulk may be destroyed, etc. Despite this, it was a factor that lowered reliability.

[Object of the Invention] The present invention has been made to solve the above-mentioned problems, and provides a computer with higher reliability by automatically restarting and interrupting bulk duplex operation without operator intervention. The purpose of this invention is to provide a method for managing duplication of auxiliary storage devices in a system.

[Summary of the Invention] In a computer system, the present invention monitors the operating status of a bulk while the system is in operation, and automatically removes a bulk that has been disconnected from the system without operator intervention when the bulk duplication operation is interrupted. connect to the system and bulk 2
The auxiliary storage device 2 of the computer system has high reliability by restarting duplex operation or disconnecting the bulk from the system in response to a bulk failure during duplex operation.
The aim is to provide a method for managing the problem of overgrowth.

[Embodiment 1 of the Invention] [Configuration of the Embodiment] Embodiments of the present invention will be described below with reference to the drawings.

FIG. 1 is an example of a computer system configuration diagram when the present invention is applied. 1 is a CPU, 2 and 3 are bulk units, 4 is a bulk status detection unit, and a bulk ready (Read)
y ) and non-ready (Not Ready) states. To implement this function, a bulk Ready signal is input through a process input/output device (equipped with an interrupt function). 5 is a bulk operation monitoring unit, and when an input/output error occurs (an error interrupt from the OS), the bulk status detection unit 4 detects the bulk R (!ady or Not Ready).
It is activated when a state is detected, and executes the steps shown in FIG. 2, which will be described later. Reference numeral 6 denotes a bulk duplication management section, in which bulks 2 and 3 form a duplication as shown.

FIG. 2 is a flowchart showing the process flow of the bulk operation monitoring section 5, and FIG. 3 is a flowchart showing the process flow of the bulk duplication management section 6.

First, in FIG. 2, in step 821, the current bulk state is stored. In step 322, it is determined whether the bulk status is duplexing suspension or resumption, and if the duplication is suspended, the process moves to step S23 to provide information on the interruption, and if it is restarted, the process moves to step 824 to provide information on the restart. The bulk duplication management section 6 is notified of each.

Upon receiving this notification, the bulk duplication management section 6 executes the process shown in FIG. 3.゛First, in step 831, it is determined whether the notification information is interrupted or resumed, and if the notification information is interrupted, steps 832 and S33 are executed, and in the case of restart, steps 334 to 3 are executed when the system is initialized.
Steps 8310 and 5311 are executed when the system is in operation in the case of restart.

In step 832, based on the interruption information, the inaccessible bulk is recognized as a slave and the other bulk is recognized as a master,
Set the date and time of interruption in the management section of the master side. This date and time is used at the time of system initialization to determine whether bulk duplex operation is possible or whether to use the bulk as the master. Then, in the next step 833, it is recognized that the slave bulk has been disconnected from the system, and the bulk duplication operation is interrupted.

On the other hand, step 834 is for determining whether to restart the system during system initialization or while the system is in operation. Since the master and slave have not been determined at the time of system initialization, the process moves to step S35 and the set date is read from the management units of both bulks. Next, in step 83G, it is determined whether the read dates are the same. That is, it is determined whether the contents of both bulks are the same.

If the dates are different, in step 337 the bulk with the new date is recognized as the master and the other as the slave, and in step 338 the contents are copied from the master bulk to the slave bulk to match them.

If the dates are the same, one is simply recognized as the master and the other is implicitly recognized as the slave in step S39.

If the master and slave have already been determined in step 334, that is, in the case of a bulk duplex operation restart notification while the system is in operation, the slave side bulk is connected to the system in step 5310, and writing is performed. Set to permission state. Next, in step 5311, copying is performed from the master bulk to the slave bulk to match the contents.

As described above, the bulk duplex operation is restarted in steps 834, 5310, and 5311.

Note that if a factor that updates the contents of the bulk occurs during the copying in steps 838 and 5311, the contents of the slave bulk will be updated at the same time as the master bulk because the slave bulk has already been connected to the system in the previous stage. Updated by dual write.

Therefore, when the copying is completed, the bulk contents of the master and slave can be completely matched.

In the above embodiment, even if one of the bulks fails during bulk 2M operation, it is possible to automatically connect to the system and restart bulk duplex operation once the failure is restored.
There is no need for operator intervention as in the past.

[Effects of the Invention] As explained above, according to the present invention, by automatically performing bulk duplication management, it is possible to provide a highly reliable auxiliary storage device duplication management method for a computer system. I can do it.

[Brief explanation of drawings]

Fig. 1 is a configuration diagram of a computer system showing an embodiment of the present invention, Fig. 2 is a flowchart illustrating processing for monitoring the operating status of bulk, and Fig. 3 is illustrating processing for managing redundant bulk. This is a flowchart. 1...Central processing unit (CPU) 2.3...Auxiliary note 1fl device (bulk) 4...Auxiliary storage device status detection unit (bulk status detection unit) 5...Urasuke storage device operation monitoring unit (Bulk operation monitoring department) 6... Auxiliary storage @ Duplex management department (Bulk dual city management department)

Claims (1)

[Claims] One or more central processing units (hereinafter referred to as CPU)
and a duplicated auxiliary storage device (hereinafter referred to as bulk), and each duplicated bulk has two
Equipped with a function to detect whether each of the duplicated bulks is abnormal or has recovered, and when one of the duplicated bulks detects an abnormality, access to that bulk is interrupted. After that, the operation is performed only with the other healthy bulk, and when it is detected that the abnormal bulk has been restored, the contents of the other healthy bulk are copied to the recovered bulk, and then the bulk is restarted. Auxiliary storage device 2 for a computer system characterized by having a function of restarting duplex operation of
Heavy weight management method.