JPH01217656A - Initial program loading system - Google Patents

Abstract

PURPOSE:To secure the normal operation of a data processing system by using an effective IPL device number shown by an IPL device number valid/invalid setting and storing means to load in initial program among those IPL device numbers which are designated by an IPL device number setting and storing means at the time restarting system operation, etc. CONSTITUTION:The validity of the IPL device numbers designated by the IPL device number setting and storing means 8 and 9 are displayed by the IPL device number valid/invalid setting memory means 10 and 11. In case the system resident volume is separated from a system due to the fault of the system resident volume, the contents of both means 10 and 11 are invalidated by an IPL device number invalidating means with respect to the corresponding IPL device number. Then a valid IPL device number displayed by the means 10 or 11 is used to load an initial program among those IPL device numbers designated by both means 8 and 9 at the time of restarting the system operation, etc. Thus it is possible to secure the normal operation of a data processing system with respect to the fault of the system resident volume, etc.

Description

DETAILED DESCRIPTION OF THE INVENTION (Industrial Application) The present invention relates to a method for performing an initial program load (IPL) from a system resident volume in a data processing system equipped with a system resident volume.

[Conventional technology]

For example, in Japanese Patent Application Laid-Open No. 59-1991, a method was proposed that enables initial program loading in a data processing system equipped with this type of system-resident volume, even if a failure occurs in a pre-specified access path. This has already been proposed in Publication No.-3610.

In other words, in the above method, the number of IPL machine number acquisition trials corresponding to the first and second access paths is set in advance, and if the initial program load using the first access path fails, the same number of attempts will be made. The initial program load is repeated as many times as the number of retries using the access path until it succeeds, and if it still fails, the initial program load is performed using the second access path, and if the initial program load using this second access path also fails, the same access is performed. The initial program load is repeated by the number of retries until it succeeds, and if it still fails, the program ends with an error.

[Problem to be solved by the invention]

By the way, in the conventional initial program loading method mentioned above, although it is possible to deal with failures in the access path,
It is not possible to deal with a failure of the system resident volume itself, and in order to deal with this problem, it is conceivable to simply apply the commonly used technique of duplication of system resident volumes. In this system, two system-resident volumes with the same contents written therein are connected to a central processing unit or the like via separate access paths.

However, even with this method, if one of the duplicated system resident volumes occurs due to a failure, etc., and the operating system is operating after disconnecting one of the system resident volumes, the system may be stopped for some reason. When the data processing system is subsequently restarted, information indicating whether the IPL machine number is valid or invalid is not taken into account, so the initial program may be loaded from the faulty system resident volume, which may prevent the data processing system from operating normally. There was a problem that the operation could not be guaranteed. That is, even in this case, although it is possible to deal with failures in access paths, it is not possible to deal with failures in system-resident polynemes.

The present invention has been proposed in view of the above points, and its purpose is to provide an initial program loading method that can ensure normal operation of a data processing system even in the event of a failure of the system resident volume. Our goal is to provide the following.

[Means to solve the problem]

In order to achieve the above object, the present invention provides a startup processing method for a data processing system that includes dual system resident volumes and performs an initial program load from one of the system resident volumes. IPL machine number setting storage means for specifying the IPL machine number for each access path, and IPL machine number valid/invalid setting storage means for displaying the validity of each IPI and w machine number specified by the IPL machine number setting storage means. and an IPL machine for setting the contents of the ■PL machine number valid/invalid setting storage means to invalid for the corresponding IPL g number when the system resident volume is separated from the system due to the occurrence of a failure of the system resident volume or the like. IP number invalidating means, and the IP number setting storage means specifies
Among the L machine numbers, the valid IPL machine number displayed by the IPL machine number valid/invalid setting recording means is used to perform initial program loading.

[Effect]

In the initial program loading method of the present invention, the IPL
The validity of each IPL machine number specified by the machine number setting storage means is displayed by the IPL machine number valid/invalid setting storage means, and when the system resident volume is separated from the system due to a failure of the system resident volume, etc. Then, the content of the iPL machine number valid/invalid setting storage means is set to invalid for the corresponding PL machine number by the IPL machine number invalidation means, and when system operation is resumed, etc., the IPL machine number setting storage means instructs the IPL Among the machine numbers, the initial program is loaded using the valid 1PL machine number displayed by the IPL machine number validity/invalidity setting storage means.

〔Example〕

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

FIG. 1 is a block diagram showing an embodiment of a data processing system to which the initial program loading method of the present invention is applied. In FIG. 1, a central processing bag R (CPU) 1 includes a channel device (CH) 2-1.2-2 and a disk control device (D
KC) 3-1.3-2 and a magnetic disk device (Disk
) 4-1 to 4-4 are connected to form two access paths, and a magnetic disk unit W4-1 and 4-2 have system resident volumes with the same contents mounted.
Initial program loading is possible from either of the following. Also, during normal operation, the magnetic disk device 4-1.
The same content is always maintained in 4-2 by writing simultaneously, and even if a failure occurs in magnetic disk device 4-1, operation can continue with only magnetic disk device 4-2. It looks like this.

On the other hand, a service processor (SVP) 5 connected to the central processing unit 1 constitutes one subsystem independent of the central processing unit 1, and monitors, operates, and controls the system. .

FIG. 2 shows an example of the configuration of the service processor 5, which controls the IPL machine number setting register 8.9, the PL machine number validity display flag 10.11, and performs initial program loading. It has an IPL processing section 12, and further includes a display console 6 and a floppy disk device 7.
are now connected.

Here, the IPL machine number setting register 8.9 is for setting the IPL machine number that specifies the access path for initial program loading, and the PL machine number setting register 8 is for setting the magnetic disk device. r! ! , 4-1 is set, and the IPL n number setting register 9 contains the magnetic disk device! The IPL machine number that specifies the access path related to 4-2 is set. The setting of these ■PL machine numbers to the IPL machine number setting register 8.9 is performed by inputting a command from the display console 6, etc.

In addition, the IPL machine number validity display flags 10.11 are the IP addresses set in the IPLQ number setting registers 8.9 and 10.11, respectively.
This is to display the state of the L machine's mobility, that is, whether it can be used without any problems or the like. In addition, these IPL8! number validity display flag 10
．． 11, the settings 5 can be changed by commands from the display console 6, and furthermore, the IPL processing section 12
Settings and changes can also be made by instructions from an operating system (not shown) running on the central processing unit 1. In addition, IPL machine number setting registers 8, 9.
The contents of the IPL machine number validity display flag 10゜11 are stored in the floppy disk device 7, and if there is no need to change them, they are read out from the floppy disk device 7 and set as IPLW.
Machine number setting register 8.9. The IPL machine number validity display flag is set to 10.11.

Hereinafter, the operation of the above embodiment will be explained along the flowchart of FIG.

First, when a request to load an initial program is issued to the service processor 5 from the display console 6 or the like, the IPL processing section 12 is activated. IPL processing section 12
First, check whether the IPL machine number #l held by the IPL machine number setting register 8 is valid or not by referring to the contents of the IPL machine number validity dynamic flag IO (Step ■), and if it is valid, the IPL machine number Load the initial program from number #1 (step ■), and if invalid, proceed to processing related to IPL machine number setting register 9 (steps ■-■). In step ■, load the initial program from IPL machine number #1. In this case, if the initial program load is successful, the IPL processing ends normally (step 7). If the initial program load from IPL machine number #1 fails, the IPL machine number validity display flag 10 Reset and disable step ■), IPL machine number setting register 9
The process moves on to processing related to (steps ■-■).

Next, regarding the IPL machine number setting register 9, it is similarly determined whether the IPL machine number #2 held therein is valid.
Make a decision by referring to the contents of the L machine number validity display flag 11 (Step ■), and if IPLa number #2 is valid, perform an initial program load from IPL machine number #2 (Step ■).
, if it is invalid, it ends with an error (step ■). When the initial program load is performed from ■PL machine number #2 in step ■, if the initial program load is successful, the IP'L process ends normally (step ■).

[phase]). Also, if the initial program load fails,
The IPLa number validity display flag 11 is reset to be invalid (step 0), and the process ends with an error (step @).

Note that when the operating system detects a failure in the magnetic disk device 4-1 or the magnetic disk device 4-2 during operation and disconnects the device from the system, or when the device is disconnected by an operator's command, the corresponding IP
Reset L machine number validity display flag 10.11.

〔Effect of the invention〕

As explained above, in the initial program loading method of the present invention, among the IPLa numbers specified by the IPL machine number setting storage means when system operation is resumed, the valid IPLM1 displayed by the IPL machine number valid/invalid setting storage means is Since the initial program load is performed using the number, the initial program load is not performed from the system resident volume that has been disconnected from the system due to a failure, etc., and the effect is that the normal operation of the data processing system can be ensured. There is.

[Brief explanation of the drawing]

FIG. 1 is a block diagram showing an example of a data processing system to which the initial program loading method of the present invention is applied; FIG. 2 is a block diagram showing an example of a service processor in FIG. 1; and FIG. 3 is a flowchart showing an example of the operation of the IPL processing section in the figure.

Claims (1)

[Scope of Claims] A startup processing method for a data processing system that includes dual system resident volumes and performs an initial program load from one of the system resident volumes, comprising: an access path for each of the system resident volumes; An IPL machine number setting storage means for specifying an IPL machine number, and each IP specified by the IPL machine number setting storage means.
an IPL machine number valid/invalid setting storage means for displaying the validity of the L machine number;
and IPL machine number invalidating means for setting the contents of the IPL machine number valid/invalid setting storage means to invalid for the corresponding IPL machine number, and among the IPL machine numbers specified by the IPL machine number setting storage means. An initial program loading method characterized in that an initial program load is performed using a valid IPL machine number displayed by the IPL machine number validity/invalidity setting storage means.