JPS62159241A - Back-up method for data processor - Google Patents

Abstract

PURPOSE:To eliminate the load of the mutual state monitor between the working and spare data processors by monitoring the answer of the working data processor through a terminal equipment against a service request and coping with abnormality of the working data processor via the spare data processor. CONSTITUTION:A host data processor 1, a spare data processor 2 and terminal equipments 3-4 are connected to a token ring 6. Thus a local area network is formed. The equipment 3 sends a service request to the processor 1 via an information frame 7 for monitor a fact that the sent frame 7 returns after having a round along the ring 6. When the returned frame 7 has no reception confirming flag of the processor 1, the equipment 3 decides the abnormality of the processor 1 and informs it to the processor 2. Thus the processor 2 checks again the processor 1 and separates the processor 1 when confirming the abnormality of the processor 1. Then the frames showing that the processor 1 is switched to the processor 2 are transmitted to all stations.

Description

DETAILED DESCRIPTION OF THE INVENTION [Field of Application of the Invention] The present invention relates to failure countermeasures in a data processing system, and in particular, to a data processing device in a data processing system in which a terminal device and a data processing device are reversely connected through a communication network. Regarding abnormality monitoring and backup.

[Background of the invention]

A spare data processing device is prepared in the data processing system so that it can be replaced at any time, and a faulty working device is immediately replaced by the spare device. The so-called hot standby method is already well known.

In the hot standby system, it has conventionally been common for the active machine and the standby machine to constantly monitor each other's status.

Recently, data processing systems formed via LAN (Local Area Network) have become popular. In many systems of this type (for example, POS systems), a data processing device having a microprocessor as a CPU is used as a host. Even in such systems, improving reliability by introducing a hot standby method is one of the important issues. However, for the 6-micro processor, constant mutual status monitoring between the active machine and the standby machine becomes an excessive burden. However, the use of a larger processor is undesirable because it increases costs. Additionally, the need for special hardware interfaces for mutual status monitoring further increases cost.

[Purpose of the invention]

An object of the present invention is to remove the burden of regular mutual status monitoring from a data processing device, and thereby facilitate the introduction of a hot standby method even in a system that uses a small processor as a CP[J]. .

[Summary of the invention]

For the terminal equipment side, if the data processing equipment goes down,
It only becomes a problem when you can no longer receive that service. From this point of view, constant mutual status monitoring between data processing devices is not necessarily essential. A feature of the present invention is that the constant mutual status monitoring between data processing devices is eliminated, and instead, the terminal device monitors the response of the data processing device (for example, a working device) to a service request issued from the terminal device. If a proper response is not obtained, use another processing device (for example, a standby device).
The abnormality is reported, and the data processing device that receives this report takes steps to deal with the abnormality (e.g. re-examination,
The main point is to switch the current machine, etc.).

[Embodiments of the invention]

FIG. 1 shows an overview of an example of a data processing system to which the present invention is applied. Two data processing devices, namely a host (CI) 1 and a standby device (C2) 2, and a plurality of terminal devices (T+, T2.T3) 3.4.5 are connected to a communication network 6. In this embodiment, the communication network 6 is a so-called token ring (a ring-shaped LAN in which access is controlled by a token passing method), and information is exchanged in the form of a frame 7.

FIG. 2 shows formats A to E of various frames 7. In these formats, S is the message start code, sMC is the message class, SA is the source address, DA is the destination address, E is the message end code, EDA is the destination address list end code, AK is the acknowledgment flag, and EAK is the receipt. Confirmation flag list end code, TN represents the transaction number, and ET represents the error type.

The type of message is indicated by the value of MC.

Token 7a usually does not have a DA designation and is a fleet
Any station can acquire it and make a transmission. However, if necessary, e.g.
When requesting processing from a specific station in order to deal with the occurrence of an abnormality, the DA can specify the destination of the token. The data frame 7b is used for normal data exchange.

This frame has an arbitrary number of destination addresses DAI to DA2.
It is possible to have reception confirmation flags AK1 to AK2 corresponding to each destination, and it is also possible to send the same data to multiple stations in one data frame and obtain reception confirmation from each destination. can.

The abnormality report frame 7c is used to send an abnormality report from the terminal device to the host or standby device.

Details of the error (for example, an error in the response from the host).

An abnormal response on one side of a duplicate file, a transaction number of data for which there is no confirmation of receipt, etc.) are entered in the ET field. The status adjustment frame 7d is used to check the status between the host and the standby device. The TN field of this frame is used to specify a particular transaction number and determine whether it has been received. The system configuration information frame 7e is a frame that is broadcast when the system configuration is changed due to a failure or the like, and the content of its "data" field is information indicating the new system configuration.

First, a case where the terminal device 3 requests a service from the host 1 will be described. FIG. 3 is a flowchart of the process. When the terminal device 3 acquires the free token, it sets the address of the host 1 as the destination address of the data 7 frame 7b.

Send (301). Each station constantly monitors the destination address field of frames on the token ring 6.

When the host device detects its own address in the destination address field of the incoming frame, it copies the frame into its buffer, sets the corresponding reception confirmation flag to '1'', and then resends the frame.

When the sent frame goes around the ring 6 and returns, the terminal device 3 checks the reception confirmation flag of the host 1 (302, 303). If the reception confirmation flag is '1'',
Host 1 is considered to be operating normally, and terminal device 3
releases the token and ends processing. On the other hand, if the reception confirmation flag is not '1', the terminal device 3 performs a certain recovery process (for example, retransmission) (304, 305), and if the recovery is successful, ends the process in the same way as above.However,
If the recovery process is not successful, the terminal device 3 sends an abnormality report frame 7C to the ET indicating an abnormality in the host's response.
It is set to a value and sent to the backup device 2 (306).

When the standby device 2 receives the abnormality report frame, 7c,
Checking the ET of that frame, it learns that an abnormal response from host 1 has been reported, 9 (307), sends a status adjustment frame 7d to host 1 (308), and checks host 1's response to it.・Ru (309). In the simplest case, the state intermediary frame 7d sets its acknowledgment flag ′
If it returns after leaving it on, it can be determined that there is an abnormality in host 1. Alternatively, internal state information sent from host l may be checked. If no abnormality is found in the host 1 as a result of the investigation, the standby device 2 notifies the host 1 that the abnormality report has been received from the terminal device 3 (310).
Finish the process. Thereafter, the host 1 only has to take necessary measures (check, disconnect, etc.) with respect to the terminal device that has issued the abnormality report.

On the other hand, if an abnormality is found in the host 1 as a result of the investigation, the standby device 2 disconnects the host 1 (311).

The system configuration information frame 7e indicating that the host has been switched to the standby device 2 is broadcast to all stations (312'), the token is released, and the process ends.

Next, a case will be described in which the host 1 and the backup device 2 form a dual file system, and the terminal device 3 synchronizes and updates the files of both. FIG. 4 is a flowchart of the process. The terminal device 3 sends the data frame 7b to both the host 1 and the standby device 2 (401), -
Check the reception confirmation flag of the circumference data frame (4
02.403).

If the reception confirmation flags corresponding to the host 1 and the standby device 2 are both '1', the synchronous update of both files is considered successful, the terminal device 3 releases the token, and ends the process. If both flags are uO'', host 1
It is determined that there is an abnormality in both the terminal device 2 and the standby device 2, and the terminal device 3 thereafter operates in offline mode (404). However, if only one of the flags is °′0″,
The terminal device 3 reports the abnormality of the other device to the device corresponding to the flag having the value "1" (405).

The device that received the abnormality report (406) sends a status adjustment frame to the other device (407) and examines the response (408), as in the case of FIG. the result,
If it is determined that there is no abnormality, the terminal device 3 is instructed to retry (409).

On the other hand, if it is determined that there is an abnormality in the other device as a result of the investigation, that device is disconnected (410), and then a system configuration information frame 7e indicating the new system configuration is broadcast (411). . At this time,
If the device that received the abnormality report is the standby device 2, the host is replaced.

In the second example, when the terminal device detects an abnormality in the response, instead of immediately reporting the abnormality.

You may try again. In such a modification, if no abnormality is found in the re-examination in response to the abnormality report, the host checks the terminal device.

In a system where multiple spare machines are prepared, these spare machines are ordered in advance in terms of acquiring control rights when an abnormality occurs, and reports of host abnormalities are sent to the spare machine with the highest order of acquiring control rights. be exposed. communication network 6
may be of a type in which each station has equal access to transmission rights1, and is not limited to token ring.

In the embodiment described above, since the status inquiry between the host and the standby device is carried out via the communication network,
No special hardware interface is required for status monitoring.

〔Effect of the invention〕

According to the present invention, each data processing device is freed from the burden of constantly monitoring each other's status. Therefore, even in a system in which an extremely small data processing device is used as a host, reliability can be easily improved by adopting the hot standby method.

[Brief explanation of drawings]

FIG. 1 is a block diagram showing an example of a data processing system to which the present invention is applied, FIG. 2 is a format diagram of various frames used for exchanging information in the system of FIG. 1, and FIGS. 3 and 4 are 3 is a flowchart illustrating an example of a backup process according to the present invention. 1...Host, 2...Spare machine, 3-5...Terminal device, 6...Communication network (token, 7,
To 7a~7e...Frame, 301, 401...Service request step from terminal device, 302~3o3°402
~403...Response monitoring step 77' in the terminal device,
306, 405... Abnormality reporting step, 308-31
2,407-411... Abnormality handling processing step.

Claims (1)

[Claims] 1. In a data processing system in which at least one terminal device and a plurality of data processing devices are connected to a communication network, the step of sending a service request from the terminal device to at least one of the data processing devices. and a step of monitoring the response of the destination data processing device to the service request sent by the terminal device, and when an abnormality is found in the response of one data processing device, the terminal device informs the other data processing device of the abnormality. A method for backing up a data processing device, the method comprising: reporting the abnormality; and performing processing to deal with the abnormality in the data processing device that has received the report. 2. In claim 1, one of the plurality of data processing devices functions as a host during normal times, and the others function as backup devices ordered in order to acquire control rights when an abnormality occurs. Method. 3. In claim 2, the service request step sends a service request to the host and at least one standby device, and the abnormality reporting step includes sending a service request to the host and at least one standby device, and in the abnormality reporting step, if the host has an abnormality, A method for backing up a data processing device that reports to a high reserve machine, and reports to the host if there is an abnormality in the reserve machine that is the destination of the service request. 4. In Claims 1, 2, or 3, the abnormality handling step includes a step of checking the state of the data processing device in which the abnormality has been reported via the communication network, and processing the data determined to be abnormal as a result. A data processing device backup method comprising the steps of disconnecting the device and reconfiguring the system. 5. The data processing device backup method according to claim 4, wherein the communication network is a token ring.