JPS626333A - Data processing system - Google Patents

Abstract

PURPOSE:To ensure effective application of a processor by dividing a service processor for maintenance/diagnosis control into two classes and separating the process synchronous with a data processing subsystem from other processes to share the operation individually. CONSTITUTION:When a central processing unit 12 has a trouble, the working of the unit 12 is discontinued and the generation of this trouble is informed to the 1st service processor 21. The processor 21 reads out the internal informa tion on the unit 12 through a scan path and stores it in a storage device 20. If it is decided that a retrial is possible with said internal information on the unit 12, the processor 21 starts the clock 12 of the unit 12 to execute the retrial. When this retrial succeeds with no trouble produced from the unit 12, the proces sor 21 informs the generation of the trouble and the success of the retrial to the 2nd service processor 30. The processor 30 receives the trouble information from the processor 21 and stores it in a disk device 32 in the form of the mainte nance information.

Description

DETAILED DESCRIPTION OF THE INVENTION (Field of Industrial Application) The present invention relates to a data processing system, and more particularly to the configuration of a maintenance/diagnosis processor that controls processing, maintenance, and diagnosis at the time of failure of each device constituting the system.

(Prior Art) Service processors have traditionally been employed in data processing systems as processors dedicated to maintenance and diagnosis. Such a service processor is used to perform failure processing when a failure occurs in the data processing device, processing for maintenance of the pseudo panel function, and diagnostic processing in controlling the execution of a diagnostic program. In addition, such a processor operates independently of a main body system such as a central processing unit in a data processing system, and is itself a computing unit, a main storage device, and an external storage device such as a disk device.
It is configured as an independent processor with a console or printer device.

Generally, the service processor is realized by a processor whose performance is lower than that of the main system equipment (hereinafter, this group of equipment is referred to as the data processing subsystem) that receives maintenance and diagnostic services from the service processor. The same processor is commonly used for various data processing subsystems ranging from high performance to low performance.

On the other hand, the following method is adopted for diagnosing the data processing subsystem and its subordinate input/output devices.

In other words, for diagnosis of each device that makes up the data processing subsystem, there is a micro-diagnosis method that uses the firmware function of each device, and FLT (FLT) that uses a scan path.
Fault Locating Meeting Test: Fault
There are locating 'l'est@new methods and software diagnostic methods that employ software instructions.

Micro diagnosis and FLT diagnosis usually use a method in which diagnostic data is loaded from a service processor and executed.In the software diagnosis method, diagnostic data is loaded from an input/output device under the data processing subsystem independently of the service processor. A method of loading is adopted.

When diagnosing input/output devices under the data processing subsystem,
A diagnostic program for the input/output device to be diagnosed is running on the data processing subsystem. In addition to supporting the above-mentioned diagnosis, the service processor has a pseudo-panel function for maintenance of the data processing subsystem, an initial setting function when starting up the data processing subsystem, and a function for providing support when a failure occurs in a device within the data processing subsystem. It supports failure processing functions such as failure information logout function and retry processing and restart processing.

(Problems to be Solved by the Invention) In the data processing device using the conventional service processor described above, as the data processing subsystem becomes more high-performance and larger, there is a difference in performance between the two in maintenance diagnosis processing. However, there is a problem in that it is not possible to perform maintenance diagnosis processing closely related to the operation of the data processing subsystem.

For example, when a failure occurs in a device that constitutes a data processing subsystem, the failure is often an intermittent failure, and retry processing is generally performed to recover from the failure. This process includes logging out (collecting) the status at the time of failure, determining whether retry is possible, resetting the environment to allow retry if retry is possible, checking retry results, and operating retry results. Frequent processing such as notification to the system is required. However, since the operation is temporarily stopped during the above-mentioned processing period, there is a problem in that if the performance of the service processor is low, the processing time becomes long and the system operation is hindered.

When executing the diagnosis of the data processing subsystem and its subordinate input/output devices, the monitor functions necessary for controlling the diagnosis execution and checking the diagnosis results are not executed on the device being diagnosed, but on other processors. should be executed. However, in the execution of software diagnosis using the conventional technology, there is a drawback that it is difficult to realize the maintenance diagnosis control that should originally be realized, such as executing the above-mentioned functions on the device to be diagnosed.

An object of the present invention is to provide two service processors for diagnosis control in a layered manner as processors that execute maintenance diagnosis of a data processing subsystem, and to have functions that are closely interfaced with the data processing subsystem and to be processed synchronously. In this system, the diagnostic functions that handle a large amount of data are assigned to the first service processor, and the other processing and the interface processing with maintenance personnel and operators are assigned to the second service processor. The object of the present invention is to provide a data processing system configured to eliminate the above-mentioned drawbacks and realize the above-mentioned service functions on a device to be diagnosed.

(Means for Solving the Problems) A data processing system according to the present invention includes a data processing subsystem, a first service processor, a second service processor, and a storage device.

The data processing subsystem is composed of at least one main storage device, main storage control device, central processing unit, and input/output control device, and is for executing data processing.

The first service processor is connected to each device constituting the data processing subsystem via a diagnostic interface, and is for managing maintenance diagnostic control processing.

The second service processor is connected to the first service processor and is used to share a part of maintenance diagnosis control processing under the first service processor.

The storage device is for storing diagnostic programs executed on the first and second service processors and diagnostic data for the data processing subsystem.

In the above configuration, the present invention accesses the storage device through the second service processor, and processes maintenance diagnosis control processing of the data processing subsystem according to predetermined assignments by the first and second service processors. It is configured so that it can be done.

(Example) Next, one embodiment of the present invention will be explained with reference to the drawings.

FIG. 1 is a block diagram showing one embodiment of a data processing system according to the present invention. In FIG. 1, 1 is a data processing subsystem, 10 is a main storage device, 11 is a main storage controller, 12 is a central processing unit, 13 is an input/output controller, 2G is a storage device, and 21 is a first service processor. ,3
o is the second service pro, 31 is the C-RT/keyboard device, black! is a disk device, and sa is a printer device.

In FIG. 1, a data processing subsystem 1 includes a main storage device 10, a main storage control device 11, a central processing unit 12, and an input/output control device 1s. The central processing unit 12 accesses and executes programs stored in the main storage device 10 via the main storage control device 11. The input/output control device 15 receives input/output instructions from the central processing unit 12,
Or, the main storage device 10 and the input/output device (
(not shown).

It has the following two processors as processors that perform maintenance diagnosis of the data processing subsystem. The first service processor 21 includes a main storage device 10, a main storage control device 11, and a central processing unit 1 that constitute the data processing subsystem 1.
2 and the signal lines 103 to 1 to the input/output control device IS.
06, and sequentially reads and executes programs stored in the storage device 20 independently of the data processing subsystem 1. Second service processor! 0 also operates independently of the data processing subsystem 1 and the first service processor 21 and performs maintenance/diagnosis control, and has a CRT/keyboard device S1 and a disk device S1 under it.
2, and input/output devices such as a printer device 35. The first service processor 21 and the second service processor 30 are connected via a signal line 108, and a communication function between programs executed on each processor is realized.

Both service processors 21 and SO normally operate in an equal relationship with each other, but perform special control regarding access to the disk device s2 as described below. Disk device 3
The file structure on 2 is an area logically managed by the first service processor 21. -1@2 service processor 30 and an area managed by service processor 30, and access management is performed from both service processors 21 and 30. The disk access control method of both processors 21°sO is such that the second service processor 30 directly starts an input/output program for the disk device 32, thereby accessing the disk device a2. When the first service processor 21 accesses the disk device 52, the input/output program to be accessed is assembled by the program on the first service processor 21, and the second service processor 3
The second service processor 30 obtains the input/output program and controls the disk device 52, and the first service processor 2
1 to the disk device 32, or control the sending of successive data from the disk device 52 to the first service processor 21. That is, when the first service processor 21 accesses the disk device 52, the second service processor BO operates as a disk control device.

Next, maintenance diagnosis control of the data processing subsystem 1 using both processors 21.2$0 will be explained.

Activation of maintenance diagnosis processing by maintenance personnel, output of processing results, and output of fault information are performed using the CRT/keyboard device 31 and printer device ss connected to the second service processor sO.

When diagnosing the data processing subsystem from the second service processor 30, the second service processor sO first performs the diagnosis on the first service processor 21 in response to a diagnosis execution instruction from the CRT/keyboard device s1. Instruct to start.

According to the above instructions, the first service processor 21 loads a diagnostic program into the storage device 20 according to the access method of the disk device 32, and connects the diagnostic signal lines 105 to 106.
to perform diagnostics on the target device. The execution result of the diagnosis is sent from the first service processor 21 to the second service processor BO, and the result is sent to the CRT/keyboard device s.
1 or output to printer device B3.

FIG. 2 is a flowchart showing internal failure processing in the data processing subsystem 1.

Next, a process to be performed when a failure occurs in a device inside the data processing subsystem 1 will be described with reference to FIG. 2.

When a failure occurs in the central processing unit 12, the clock of the device is stopped and the first service processor 21 is notified of the failure. In response to this notification, the first service processor 21 reads the internal information stored in the registers, clip-flops, and failure flags of the central processing unit 12 through the signal line 105 and stores it in the storage device 20. Store.

Next, the service processor 21 of aEl determines whether retry is possible based on the above internal information, and if retry is possible, changes the internal state of central processing unit 12 to a state where retry is possible. In order to do this, the previously read internal information is edited and embedded in the scan path.

Furthermore, the first service processor 21 is the central processing unit 1
Start the second clock and make it retry. When executing the above retry, no failure occurs from the central processing unit 12,
Upon recognizing that the retry process was successful, the first service processor 21 notifies the second service processor 30 of the occurrence of the failure and the success of the retry process. Based on this notification, the second service processor 30 obtains the fault information from the first service processor 21, and
This information is also stored as maintenance information.

In the above embodiment, the division of processing between the first service processor 21 and the second service processor 30 has been explained. The service processor 21 and the m2cot-bis processor 30 share and execute the processing.

Although the embodiment shown in FIG. 1 shows the case where each device constituting the data processing subsystem 1 is one, maintenance functions and diagnostic functions are similarly provided in the case of a configuration with two or more multi-processors. can be supported.

(Effects of the Invention) As explained above, the present invention provides two service processors for diagnosis control in a hierarchical manner as processors that execute maintenance diagnosis of the data processing subsystem, and has a close interface with the data processing subsystem. Functions that must be processed synchronously and diagnostic functions that handle large amounts of data are assigned to the first service processor, and other processing and interface processing with maintenance personnel and operators are transferred to the second service processor. By sharing the functions, it is possible to take advantage of the advantages of each processor, and the necessary maintenance and diagnosis functions can be efficiently realized.

In addition, a low-speed and low-cost second service processor is commonly provided for various models of data processing subsystems with different performance, and a first service processor is installed individually for each data processing subsystem. By providing
This has the effect of making the entire system common and realizing a diagnostic control processor that can satisfy the trade-off between performance and price.

[Brief explanation of the drawing]

FIG. 1 is a block diagram showing one embodiment of a data processing system according to the present invention. FIG. 2 is a flowchart showing an example of processing when a failure occurs in the central processing unit in FIG. 111...Data processing subsystem 10...Main storage device 11...Main memory control device 12...Central processing unit 15...Input/output control device 20...Storage device 21.30...Life service Processor 31 --- CRT/keyboard device 5.2 --- Disk device 33 --- Printer device 100~108.110~112.200~2On...
···Signal line

Claims (1)

[Claims] a data processing subsystem for executing data processing, each consisting of at least one main storage device, a main storage control device, a central processing unit, and an input/output control device, and the data processing subsystem; a first service processor connected to each of the devices via a diagnostic interface and for managing maintenance diagnosis control processing; and a first service processor connected to the first service processor and subordinate to the first service processor. a second service processor for sharing a part of the maintenance diagnosis control processing; and a storage for storing diagnostic programs executed on the first and second service processors and diagnostic data of the data processing subsystem. a storage device, accessing the storage device via the second service processor, and predetermining the maintenance diagnosis control processing of the data processing subsystem by the first and second service processors. 1. A data processing system characterized in that the data processing system is configured to perform processing according to assigned assignments.