JPS62190543A - Control system for quick restoration from trouble of communication system - Google Patents

Abstract

PURPOSE:To quickly restart the business and gather NCP (communication control program) dump by switching a stand by CCP (communication controller) to a current NCP to restart the business and gathering dump information related to a faulty NCP when the current NCP is faulty. CONSTITUTION:When an NCP trouble detecting part 16 detects the trouble of an NCP14 operated on a current CCP12, a CCP switching control part 20 is started. The control part 20 searches a corresponding stand-by CCP13 in accordance with a CCP control table 21 and switches this CCP13 to a new current CCP. A business restart part 22 causes the CCP13 to perform data communication, which is performed by the CCP12 before the occurrence of the trouble, and activates an NCP15 to execute the restart processing of the business. In parallel with this business restart, a dump gathering part 23 executes the processing to output memory information related to the faulty NCP14 to a prescribed dump output device 24.

Description

[Detailed Description of the Invention] [Summary] When the communication control program (NCP) installed in a communication control device (CCP) goes down, another CCP is secured, and the business is immediately restarted using that CCP, and
For the failed NCP, after this switchover is completed,
By collecting a dump, the system can be started up quickly after the NCP goes down, and dump information can be obtained for troubleshooting.

[Industrial application field]

The present invention relates to a communication failure high-speed start-up control method in a system that processes data by controlling input/output related to data communication using a communication control program installed in two communication control devices.

[Conventional technology]

FIG. 5 is a diagram for explaining an example of the conventional method.

A failure occurred in the communication control program (NCP) 14 installed in the communication control device (CCP) 12.

When the NCP 14 goes down, in the case of the conventional method, the host processing unit 10 first dumps the state of the NCP 14 at the time of the down time on a magnetic disk device, etc. in order to obtain information necessary for investigating the cause of the failure such as a software bug. Output device 24
Output to. Thereafter, the NCPL4 is reloaded into the CCP 12 from the file in which the NCP load module 18 is stored, the NCP 14 is activated, and then the work interrupted due to the down is resumed.

[Problem that the invention seeks to solve]

According to the above-mentioned conventional method, since a long processing time is required to collect the NCP dump, there is a problem that business operations cannot be continued in a short time. especially.

Problem 9 is a big problem in online systems, such as those related to finance, which require fast start-up when down.

On the other hand, if you omit collecting dump information in order to start up quickly, if the cause is a software bug, there will be insufficient information to detect that bug, making it impossible to debug, and if the same cause occurs again. The problem is that it can sometimes go down.

SUMMARY OF THE INVENTION The present invention aims to solve the above-mentioned problems and to provide a system that enables quick resumption of business operations and that can collect NCP dumps necessary for fault investigation.

[Means for solving problems]

FIG. 1 shows an example of the basic configuration of the present invention.

In FIG. 1, 10 is a post processing device that processes data, 11 and 12 are current communication control devices (CCP),
13 is a backup CCP, 14 and 15 are communication control programs (NCPs) that control the CCP, 16 is an NCP failure detection unit that detects abnormalities in the NCP, and 17 is a c that manages the CCP.
cp management department.

18 is an NCP load module; 19 is a preloading control unit that controls loading the NCP to the backup CCP 13 in advance; and 20 is a CCCP that performs data communication for business purposes.
CCP switching control unit for switching between active and standby, 2
Reference numeral 1 represents a CCP management table, 22 represents a business resuming unit that restarts a business interrupted due to a failure, 23 represents a dump collection unit that collects a dump about the failed NCP, and 24 represents a dump output device that outputs dump information.

In response to a request from the host processing device 10, the CCP that performs data communication control is 1 active CCP II used for normal operation,
12. . . , and a backup CCP 13 that can be used as a substitute when the working CCP fails are provided in groups. Correspondence information for the active CCP and backup CCP is as follows:
It is managed by the ccp management table 21. Also, c
The cp management table 21 contains the NCP that operates the cp.
Information on whether or not is loaded is also stored.

The preloading control unit 19 activates the NCP 14 such as the current CCP i2 in response to an activation (ACT) instruction from the NCP or a preloading instruction from the operator.
The same NCP load module 18 as the spare CCP 13
It executes the loading process for the file.

For example, when the NCP failure detection unit 16 detects that the NCP 14 running on the active CCP 12 is down, the ccp
The switching control section 20 is activated. ccp switching control unit 20
is the corresponding backup CC according to the ccp management table 21.
P13 is found and the spare CCP13 is newly switched to the current one.

The business resumption unit 22 uses the backup CCP 13 to perform data communication that was being performed by the active CCP 12 before the failure.

It takes over the task, activates the NCP 15, and executes processing to restart the business.

In parallel with the restart of the business by the business restart unit 22, the dump collection unit 23 executes a process of outputting memory information related to the down NCP 14 to a predetermined dump output device 24.

[Effect]

According to the present invention, when the current NCP 14 goes down, the state of the memory, etc. of the current CCP 12 is left as is,
By the CCP switching control section 20.

Immediately switch to backup CCP 13 and restart business department 2.
2, the business is restarted.

Therefore, start-up control can be performed quickly.

On the other hand, after switching to the standby mode, the dump collection unit 23
Original working CCP1 with failure occurrence N(1)14
2 and collects the dump of that NCP 14, this collection process can be executed without affecting business processing, and the cause of the failure can be searched for at any time using the information stored in the dump output device 24. be able to.

〔Example〕

FIG. 2 is an explanatory diagram of control according to an embodiment of the present invention, FIG. 3 is an explanatory diagram of preloading processing, and FIG. 4 is an explanatory diagram of processing when a failure occurs.

As shown in FIG. 2 (al), the host processing device'io is

Activation instruction to NCP 14 in advance or reserve C
If there is a preloading instruction specifying CP 13,
Load the NCP 15 and make it standby. Then, when the NCP 14 of the current CCP 12 goes down due to a failure,
Switch the active CCP 12 to the backup CCP 13 (Fig. 2 T
b), (C1). Immediately thereafter, the NCP 15 restarts the business (FIG. 2(d)), and a dump is collected offline for the faulty NCP 14, regardless of business processing.

Next, the processing by the preloading control section 19 shown in FIG. 1 will be explained with reference to FIG. 3. Processing numbers ■ to ■ in the following explanation correspond to numbers ■ to ■ shown in FIG.

(2) When there is an activation instruction for a working CCP, apart from the processing for the working CCP, the CCP management table 21 is searched to find a backup CCP corresponding to the working CCP. Note that preloading may be performed for a designated CCP by another operator/frmant.

■ Allocate spare or designated CCPs;
An NCP with the same content as the current CCP is loaded and placed on standby.

■ Spare or designated C of CCP management table 21
Set the preloaded display flag in the CP entry.

With the above steps, preparations for speeding up loading are completed.

If a failure occurs in the current CCP, processing is executed as shown in FIG. Processing number in the following explanation ■～■
correspond to the numbers shown in FIG.

■ When a failure occurs, determine whether it is a serious failure that requires CCP switching. In the case of a simple line failure, CCP switching is not necessary, and the same general failure handling as in the past is performed.

(2) If CCP switching is necessary, refer to the CCP management table 21 to determine whether the same type of NCP has been preloaded. If preloading has been completed, control is transferred to process (2).

■ If not preloaded, the CC where the failure occurred
Allocate a spare CCP corresponding to P.

■ Load the NCP into the allocated CCP. This loading time is generally a few seconds to 1
It takes about 0 seconds.

■ Start up the new NCP and resume the work that was interrupted due to the failure.

■ Determine whether an NCP dump is necessary.

If the failure is related to NCR software, it is determined that a dump is required. CCI)/Hardware failure such as channel failure.

If the NCP is not related to the cause of the failure, it is determined that a dump is not necessary.

■ If a dump is required, save the faulty ccplI and dump the NCP. This dump collection processing can be executed in parallel with the resumed business processing.

■ In the judgment in process ■, if it is found that preloading has been completed, the loaded CC
Allocate P.

■ Attempt to contact the NCP among them.

0 If contact is made and the NCP is found to be operational, control is transferred to process (2) without loading a new NCP.

resume business.

(2) If the preloaded NCP is unable to operate for some reason, the NCP is loaded again into the backup CCP and control is transferred to process (2).

Routes ■ and ■ shown by 9-dotted lines in Figure 4.

■, ■, @), ■, ■, ■ are high-speed routes related to the present invention, and after a failure is detected, business can be resumed immediately with the minimum necessary processing. In addition, in this embodiment, not only the dump collection time can be reduced, but also the NCR loading time can be reduced until the work is resumed.

〔Effect of the invention〕

As explained above, according to the present invention, it is possible to resume work earlier, and it is also possible to collect NCP dump information necessary for fault investigation without giving up for rapid startup. become.

[Brief explanation of drawings]

Fig. 1 is an example of the basic configuration of the present invention, Fig. 2 is an explanatory diagram of control of an embodiment of the present invention, Fig. 3 is an explanatory diagram of pre-loading processing, Fig. 4 is an explanatory diagram of processing when a failure occurs, FIG. 5 shows a diagram for explaining an example of the conventional method. In the figure, 10 is the host processing device, 11 and 12 are the current C
CP', 1'3 is backup CCP, 14 and 15 are NCP
, 16 is an NCP failure detection unit, 17 is a CCP management unit, 18
19 is an NCP load module, 19 is a preloading control unit, 20 is a CCP switching control unit, 21 is a CCP management table, 22 is a business resuming unit, 23 is a dump collection unit, 24
represents a dump output device. Patent Applicant: Fujitsu Ltd. Representative Patent Attorney Hiroshi Mori 1) (1 other person) O-cho Ikimei Issouga Yusuke 11 Toshi 11 Rike Tomomo Sun Moon Zu'f
'+2 countries 9L elephant S 2 4 rows η5 meals

Claims (2)

[Claims] (1) Comprising a host processing device (10) and a plurality of communication control devices connected to the host processing device, at least one of the communication control devices (12) is equipped with a communication control program, and the communication control In a system that processes data by controlling input/output related to data communication by a program, when a failure occurs regarding the communication control program, the first communication control device (12) on which the communication control program is installed is A communication control device switching means (20) for switching to the second communication control device (13), and a communication control program installed in the second communication control device (13), a business resuming means (22) for restarting business using the first communication control device (22); 12) A dump collection means (23) for collecting dump information regarding a faulty communication control program installed in a communication system failure high-speed start-up control method. (2) The second communication control device (13) is characterized in that a communication control program of the same type as the communication control program installed in the first communication control device (12) is installed in advance. A communication system failure fast start-up control method according to claim 1.