JPH05165798A - System controlling system for two-series system - Google Patents

Abstract

PURPOSE:To constitute a sound system of the host computer of one side system and the front end processor of the other side system even in the case that a serious fault occurs in the front end processor of one side system and the host computer of the other side system. CONSTITUTION:The positioning of a first and a second front end processors 3, 6 is connected exclusively as a duplex device independent of a first and a second host computers 1, 4, and a system monitor device 7 digitizes the various faults of the first and the second host computers 1,4 and the first and the second front end processors 3, 6, etc., in accordance with a fault level, and switches and controls the normal service and the standby host computers 1, 4 or the normal service and the standby front end processors 3, 6 at the time of the occurrence and the restoration of the fault.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a supervisory control having a front-end processor (hereinafter abbreviated as FEP), etc.
The present invention relates to a system management system of a series system.

[0002]

2. Description of the Related Art As is well known, a two-series system (duplex system) is composed of two series of computers having the same processing capacity, and usually performs different processes, but one system has some abnormality. Is generated, the other system continuously performs the processing.

In the conventional duplex system, as shown in FIG. 4, the first host computer 1 is provided with the first FEP 3 via the transmission path 2, and the second host computer 4 is provided.
The second FEP 6 is connected to the
The P3 and P6 are positioned as devices subordinate to the respective host computers 1 and 4. Further, reference numeral 7 in the figure denotes a system monitoring device transmitted via a transmission line 8, and this system monitoring device 7 monitors whether or not an abnormality has occurred in each system.

Therefore, for example, when the first host computer 1 is operating in the normal mode and the second host computer 4 is operating in the standby mode, the first FEP 3 on the normal system side has CPU power cut, PI / O power cut, WDT error, parity error,
When a major failure such as a software major failure occurs, it is treated as a major failure of the first host computer 1 on the regular system side, and the regular and standby systems of the first host computer 1 and the second host computer 4 are exchanged. It was a switching factor.

[0005]

The conventional system management method of the two-series system is configured as described above, and when it is handled as a subordinate device of the host computer, even though the FEP is duplicated, for example, When a serious failure occurs in the 1FEP3 and the second host computer 4, neither the first system nor the second system is in a healthy state, and both systems are down.

The present invention has been made in order to solve the above-mentioned problems. Even if a serious failure occurs in the FEP of one system and the host computer of the other direction, the host computer of one system and the FEP of the other system. The purpose of the present invention is to obtain a system management system for a two-series system in which a sound system can be configured by.

[0007]

According to the system management method of a two-series system according to the present invention, the positioning of the first and second front end processors is exclusively performed as a duplication device independent of the first and second host computers. Connected, the system monitoring device digitizes various failures according to the failure level, such as the above-mentioned first and second host computers and each device such as the first and second front-end processors. And standby host computers or regular and standby front-end processors.

[0008]

According to the system management method of the two-series system of the present invention, the failure of the host computer occurs / recovers during the normal operation and standby operation of the host computer, and the normal-use host computer has a higher failure level than the standby-side host computer. Mode switching signal is sent from the system monitoring device when there is a fault, and when the front-end processor connected to the normal side has a higher fault level than the front-end processor connected to the standby side, the connection switching signal is sent. Is transmitted from the system monitoring device via the host computer.

[0009]

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS An embodiment of the present invention will be described below with reference to the drawings. FIG. 1 is a system configuration diagram showing an embodiment of the present invention. In FIG. 1, the same or equivalent components as in FIG. In FIG. 1, the first FEP 3 and the second FEP 6 are transmitted to the first host computer 1 via a transmission line 9, and
For the second host computer 4, the first FEP 3 and the second FEP 3
The FEP 6 is transmitted via the transmission line 10. Further, the system monitoring device 7 is provided with a first host computer 1, a second host computer 4, a first FEP 3, and a second FE.
It is connected to P6 by hard-wired wires 11, 12, and 13, respectively.

With this configuration, the first and second
The positioning of the FEPs 3 and 6 is a common device that is connected exclusively to the first and second host computers 1 and 4 as an independent duplication device.

The common equipment including the system monitoring device 7 is managed as follows. CPU of each FEP3,6
Serious failures such as power-off, PI / O power-off, WDT error, memory parity error, and software major failure are hard-wired-OR and input to the system monitoring device 7.

Some of the FEPs 3 and 6 have a medium or minor failure signal are detected by the software in the FEPs 3 and 6, and the signals from the FEPs 3 and 6 to the host computers 1 and 4 are detected.
The system monitoring device 7 is notified via the other, and some others are hard-wired 11, 12, 13 and the transmission line 9,
The system monitoring device 7 is notified via 10 or the like.

Further, the failure signals of the host computers 1 and 4 are also hard-wired 11, 1 as in FEPs 3 and 6.
2, 13 and the transmission lines 9, 10 are notified to the system monitoring device 7. In the system monitoring device 7, various failures of the respective devices such as the host computers 1 and 4 and the FEPs 3 and 6 are defined in advance as, for example, the failure occurrence of a major failure item is 4, and the failure occurrence of a medium failure item is 3. Therefore, it is digitized as shown in Table 1 of the following table (this is called the failure level).

[0014]

[Table 1]

When a plurality of failures occur, the larger level is adopted. Further, in Table 1, the failure levels “0” and “1” are regarded as the same level. The fault level “0” indicates that the system component device has no abnormality. Further, the failure level “1” indicates that there is an abnormality in the common device or the individual device that is being used simultaneously by both systems. That is, such a device abnormality is always detected in both systems. Therefore, it is not a switching factor for ensuring that the fault levels of both systems are the same. However, it is necessary to provide "0" and "1" levels in order to distinguish whether there is an abnormality in the system configuration.

The system monitoring device 7 is used for the occurrence / failure of each device.
At the time of restoration, the fault level of each device is grasped, the fault levels are compared between the respective duplicated devices, and the one with a larger numerical value is determined to be a more serious fault. When a failure occurs / recovers in the host computer 1 on the regular side during the normal / standby operation of the host computers 1 and 4, and the host computer 1 on the regular side has a higher fault level than the host computer 4 on the standby side, the system A mode switching signal is transmitted from the monitoring device 7 via the transmission lines 9 and 10, and the host computer 1 and the host computer 4
And automatic system switching (switching between normal mode and standby mode)
Is performed (see FIGS. 2A and 2B).

Also, the first side of the first and second host computers 1 and 4 which is connected to the normal side during normal and standby series operation of the host computers 1 and 4.
Failure / recovery of EP3, 1st FEP of normal side connection
When 3 has a higher fault level than the second FEP 6 connected on the standby side, the system monitoring device 7 transmits FEPs 3, 6 and connection switching signals via the system monitoring device 7 and the host computers 1, 4 via the transmission lines 9, 10. ,
Automatic connection switching of FEPs 3 and 6 is performed (see FIGS. 3A and 3B). In addition, in FIG. 2 and FIGS. 3A and 3B, the shaded portions indicate logical connection states.

Both host computers 1 and 4 and both F
It is connected to EPs 3 and 6 on the transmission line and is in the open state, but the logical connection state is exclusively established.

[0019]

As described above, according to the present invention, the first host computer is connected to the first host computer via the transmission line.
A front end processor and a second front end processor; a second host computer connected to the first front end processor and the second front end processor via another transmission path; the first host computer; A duplexer in which the first and second front-end processors are positioned independently of the first and second host computers by hard-wired connection to the host computer and the first front-end processor and the second front-end processor, respectively. And a system monitoring device exclusively connected to the system monitoring device according to various failure levels of the first and second host computers and the first and second front end processors. Digitize and use regularly when trouble occurs / recovery And the standby host computer or the normal and standby front-end processors are switched and managed, so that even if a serious failure occurs in the FEP of one system and the host computer of the other system, the host computer of the one system and the other system There is an effect that a healthy system can be constructed by the FEP of.

[Brief description of drawings]

FIG. 1 is a system configuration diagram showing an embodiment of a system management system of a two-series system according to the present invention.

2A and 2B are explanatory diagrams of normal / standby switching operation when the normal-use host computer in FIG. 1 fails.

3A and 3B are explanatory diagrams of the operation of normal / standby switching when the front-end processor on the normal side in FIG. 1 fails.

FIG. 4 is a system configuration diagram showing an example of a system management system of a conventional two-series system.

[Explanation of symbols]

1, 4 ... First and second host computers, 3, 6 ... First and second front end processors, 7 ... System monitoring device, 9, 10 ... Transmission lines.

Claims (1)

[Claims] 1. A first front-end processor and a second front-end processor connected to a first host computer via a transmission line, and another one for the first front-end processor and the second front-end processor. A second host computer connected via a transmission path, the first host computer, the second host computer, the first front-end processor, and the second
A system monitoring device that is hardwired to the front-end processor and is exclusively connected to the first and second front-end processors as a duplication device independent of the first and second host computers. This system monitoring apparatus digitizes numerical values according to various failure levels of the respective devices such as the first and second host computers and the first and second front end processors, and performs normal and standby operations at the time of failure occurrence / recovery. A system management system for a two-series system that switches and manages a host computer or regular and standby front-end processors.