JPH09138769A - System and method for software delivery - Google Patents

Abstract

PROBLEM TO BE SOLVED: To instantly recover a software so as to use the same software by providing a software recovering means instructing the replacement software with an old version before delivery at the time of judging the failure of delivery in a client machine. SOLUTION: What is described in a server-side operation setting file 17 set to the disk device 3 on the side of a server machine 1 is the name of a software, whether to recover or not and information of the name of a module to recover. At the time of the failure of delivery, a server-side operation setting file judging means 18 judges whether to recover or not corresponding to the delivered software. When the software is judged to recover, a software recovery instructing means 15 is executed. Thereby the client machine 2 executes a software recovering means 16 to copy all the modules of the software of the old version in an original space on a disk to return the software into the original state before delivery.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a software distribution system and method for distributing software from a server machine to a client machine in a network system including a plurality of client machines connected to the server machine.

[0002]

2. Description of the Related Art Generally, regarding a software distribution method in an environment where there are a plurality of machines in a network environment, when the version of the software is modified, the old version of the software is manually replaced by the new version of the software for each machine. Since it is inefficient to replace it with, the download is being performed to distribute the new version of the software from the server machine to each client machine all at once. An example of such a software distribution system is the system described in Japanese Patent Laid-Open No. 4-359345. FIG. 11 is a diagram showing the system configuration, in which 1 is a host computer, which corresponds to a server machine. Reference numeral 2 is a terminal, which corresponds to a client machine. A LAN cable 22 serves to connect the two. Reference numeral 6 denotes a software transmitting means, which serves to transmit the new version of software to be distributed to the terminal 2. On the terminal 2 side, there is software receiving means 7, which serves to receive and store the transmitted software. A disk device 3 stores distribution target software 4 on the host computer 1 side, the transmitted software is stored on the terminal 2 side, and manages the version of the software on each terminal 2. Terminal software management table
76 is stored, and the contents of the table are updated by the management table updating means 77 in each terminal. The updated contents are transmitted to the host computer 1, and the terminal software management table 76 on the host computer 1 side is displayed.
Is also updated.

Next, the operation will be described. The distribution target software 4 is designated by the host computer 1, and is transmitted to each terminal 2 by the software transmitting means 6. At each terminal 2, it is received by the software receiving means 7 and stored in the disk device 3. After storing, upon receiving a signal from the software receiving means 7, the management table updating means 77 updates the corresponding software version on the terminal software management table 76. This update result is transmitted to the host computer 1, and the terminal software management table 76 on the host computer 1 side is also updated. Here, in an actual system under a distributed processing environment, failure of software distribution to some hosts or failure of software distribution processing of some applications may adversely affect the operation of the entire system. In addition, when software is distributed while the system is operating, it may not be possible to allocate a lot of time to the software distribution itself. In that case, it is necessary to immediately recover the system when the software distribution fails. .

[0004]

Since the conventional software distribution system is configured as described above, when the software distribution fails, the distribution fails due to the terminal software management table 76 on the host computer 1 side. Although it may be retransmitted to the terminals, it may fail again in the case of a trouble such as a line on the network, which may cause a situation in which the versions of software running on all the terminals do not match. Also,
Host to match software version
It is possible to send the software of the old version from the computer 1 to restore the software of all terminals to the old version, but again, in the case of a trouble such as a line on the network, it may fail again, and There was a problem that the versions of the software running on the terminals would not match. Further, when the number of terminals 2 that failed in software distribution is smaller than the number of terminals as a whole, it may not be necessary to restore the software distributed to all terminals to the old version. There was also a problem that it was not possible to respond flexibly according to. Furthermore, if you find that the software that failed in software distribution does not actually affect the operation of the system, or only has a limited effect, you need to restore all the software to the old version after the software distribution fails. However, there is also a problem that it is not possible to flexibly respond to system requirements with a uniform method. Furthermore, even if the terminal side succeeds in receiving the old version of the software transmitted from the host computer 1, the old version may remain for a while due to the case where the terminal 2 is using the software to be distributed or due to the processing. There was also a problem that we could not deal with it if we wanted to use this software.

The present invention has been made in order to solve the above-mentioned problems. Even if the distribution from the server machine to the client machines fails, the same software can be used in all the client machines. The purpose is to enable immediate recovery. The second purpose is that the software distribution does not succeed for all client machines or all the software, and even if the distribution result is incomplete, it is possible to flexibly respond to the system request. The purpose is to be able to take. Further, a third purpose is to decide whether to replace the software of the new version with the software of the old version immediately or after a lapse of a predetermined time depending on the convenience of the client machine even when the distribution from the server machine to the client machine is successful. The purpose is to be selectable on the client machine.

[0006]

DISCLOSURE OF THE INVENTION A software distribution system and method according to the present invention is a network system including a plurality of client machines connected to a server machine, and the distribution target software on the distribution list transmitted from the server machine. Software storing means for storing the old version of the above in the client machine, software transmitting means for transmitting the new version of software from the server machine to the client machine, software receiving means for receiving the transmitted software in the client machine, Distribution result notifying means for notifying the server machine of success or failure, distribution result judging means for judging whether the distribution to each client machine has succeeded or failed based on the notified distribution result, If it is determined that the distribution has failed in at least one client machine by the method of above, the new version software already distributed from the server machine to all the client machines is replaced with the old version software before distribution. The software recovery command means for instructing, and the software recovery means for returning the software of the old version stored by the software storage means to the original state when the software recovery instruction is received.

[0007] Further, a step of storing the old version of the distribution target software on the distribution list transmitted from the server machine in the client machine, a step of transmitting the new version of the software from the server machine to the client machine, the transmitted software Receiving at the client machine,
The step of notifying the server machine of the success or failure of the reception, the step of judging whether the distribution to each client machine has succeeded or failed based on the notified distribution result, and the distribution in at least one client machine according to this judgment result When it is determined that the above has failed, a step for instructing all client machines to replace the distributed new version software with the old version software before distribution, when this command is received, or It comprises a step of restoring the stored old version of the software to the original state when the reception fails.

Further, when the server machine determines that the distribution of the new version of the software to each client machine has failed, all the clients are determined according to the ratio of the number of client machines that failed in distribution to the total number of client machines. It decides whether or not to instruct the machine to replace the distributed new version software with the pre-distributed old version software.

Further, it is possible to set, for each software, a decision as to whether or not the new version of the software sent to the client machine should be replaced with the old version of the software.

Furthermore, the step of transmitting the new version of the software from the server machine to the plurality of client machines, the step of receiving the transmitted software in the client machine, and the received new version of the software immediately becoming the old version of the software. In each client machine, it is possible to select the replacement or the replacement after a lapse of a predetermined time.

When a preset time elapses after the new version software is transmitted from the server machine to the plurality of client machines, the new version software is replaced with the old version software from the server machine to all the client machines. The command to exchange is transmitted.

Further, it is possible to set the time for replacing the new version software sent to the client with the old version software for each software.

Further, it is possible to select whether or not to save the old version software of the new version software transmitted from the server machine in the client machine.

[0014]

BEST MODE FOR CARRYING OUT THE INVENTION

Embodiment 1 FIG. FIG. 1 shows a software distribution system according to an embodiment of the present invention. In the figure, 1 serves to transmit a new version of software based on a distribution list in a server machine. Reference numeral 2 is a client machine that receives software transmitted from the server machine 1 and deploys the software so that it can be used. A disk device 3 is a disk device 3 connected to the server machine 1.
Mainly stores the software to be distributed, and the disk device 3 connected to the client machine 2 mainly stores the received software 5. Reference numeral 6 denotes a software transmitting means for transmitting a new version of software from the server machine 1 to a plurality of client machines 2 at the same time. The data to be transmitted has a distribution list at the beginning, and the distribution list includes a list of software to be distributed this time. The table is stored. Reference numeral 7 is a software receiving means for receiving the distributed software in the client machine, and 8 is a software storing means for checking the distribution list and storing the old version of the software to be distributed in another space in the disk. Since a plurality of software 9 are packaged for transmission and cannot be used in the system as they are, the plurality of software 9 are divided by the software receiving means and expanded into a predetermined disk space of the disk device 3 after being received by the software receiving means. It is a software deployment means for performing. 10 is a distribution result notifying means for notifying the server machine 1 of the success or failure of the distribution in the client machine 2, 11 is a distribution result reception for receiving the distribution result notification from the client machine 2 in the server machine 1 and recording it in the status recording file 12. Means, the client machine 2 with the information of the distribution result judgment reference file 13 and the information of the status recording file 12 having a judgment criterion that is a ratio of the number of client machines 2 that have failed in distribution to the total number of machines. Is a distribution result judging means 14 for judging whether or not the distribution has been successful. Reference numeral 15 denotes software recovery instruction means for instructing all client machines 2 from the server machine 1 to return the software to the original state before distribution when the distribution result determination means 14 determines that the distribution has failed, and 16 indicates When a software recovery command is received, the software recovery means restores the software to the original state before distribution by moving the old version of the software stored in advance by the software storage means 8 to a predetermined disk space.

As shown in FIG. 2, the server side operation setting file 17 set in the disk device 3 on the server machine 1 side includes software name (software name), recover (whether or not to recover), and modul.
Describe the information of e list (name of module to be recovered). If the distribution is unsuccessful, the server side operation setting file judging means 18 judges whether or not the software is to be recovered depending on the distributed software, and if it is judged that the software is recovered, the software recovery instruction means 15 is executed. As a result, the client machine 2 executes the software recovery means 16, copies all the modules of the software of the old version to the original space on the disk, and restores the software to the original state before distribution. On the other hand, 19 is an operation setting file on the client side. As shown in FIG. 3, this file contains software (software name),
immediately (whether or not to immediately update), s
avemodule (whether to save the old version of software when updating), module list
Describe the information of (module name to be updated). As shown in FIG. 3, if ALL is described as (software name) and y is described as to (whether or not to immediately update), when this client machine 2 receives software, the software expanding means 9 immediately Unpack the distributed software and update to the new version. When (whether or not to immediately update) is described as n, after a predetermined time elapses, the software of the client side operation setting file 19 causes the software expanding means 9 to expand the distributed software to develop a new version. To update. Further, in the case of system management, when a separately preset time has elapsed, the software deployment instruction means 21 in the server machine 1 forcibly deploys the software distributed to all the client machines 2. 22 is the server machine 1
Is a LAN for connecting the client machine 2 to the client machine 2, and usually several to several tens of client machines 2 are connected.

Next, the operation will be described. FIG. 4 shows an embodiment of the operation of the software distribution system. First, by executing the software transmission means 6, the software is packaged from the server machine 1 to all client machines 2 in the network system. Is transmitted (step 23), and each client machine 2 receives the software by the software receiving means 7 (step 24). Depending on the command end code of the software receiving means 7, the software receiving means 7
Is succeeded (step 25), and if the software is successfully received, the distributed new version of the software is copied and stored on the disk device 3 (step 26). Next, when the same software as the received software exists on the disk device 3, the old version software currently on the disk device 3 is copied to another space on the disk device 3 by the software storage unit 8. And save (step
27). If the software reception fails, set "failure" in the software distribution result notification (step
28). Next, each client machine 2 executes the software expanding means 9 to expand the packaged new version of the software (step
29). If the software deployment is successful (step 30), the software distribution result notification is set to "success" (step 31), and if the software distribution is unsuccessful, the distribution result notification is set to "failure" (step 28). Then, each client machine 2 transmits the distribution result to the server machine 1 by the distribution result notifying means 10 (step 32). The contents of the distribution result notification are shown in FIG. The result (distribution result) is an example in the case of OK. The server machine 1 receives the distribution result notification from the client machine 2 by the distribution result receiving means 11
Is received (step 33) and the distribution result is recorded in the status recording file 12 shown in FIG. 6 (step 34). The state recording file 12 is created for each distributed software. When the reception of the distribution result notifications from all the client machines 2 is completed (step 35), in the server machine 1, the contents of the state recording file 12 are analyzed by the distribution result judging means 14 (step 36), and even one machine can be distributed. If it is determined that there is a failed client machine 2, the software recovery command means 15 commands all client machines 2 to recover the software (step 37). When the distribution is successful in all the client machines 2, the process ends as it is.

In step 35, even if the distribution results from all the client machines 2 are not received, in step 38, the elapsed time from the transmission of the new version software (step 23) is measured, After the lapse of a predetermined time, it is determined that the distribution result cannot be received due to a trouble on the network, and the software distribution itself in the client machine 2 that cannot receive is also presumed to be “failed”. 2 is instructed to recover the software (step 39). In each client machine 2 that has received the software recovery command (step 40), the software recovery means
According to 16, the old version of the software saved in step 27 is copied to the original space on the disk device 3, and the software is returned to the original state before distribution (step 41).

In this way, if any one of the client machines 2 in the network system has failed to be distributed, the software of all client machines 2 is forcibly returned to the version before distribution. Since the same version of software operates on all client machines 2 in the network system, in a system where it is strictly required that the same version of software always operates,
First aid to revert to the old version is essential and highly effective. After returning to the state before distribution, the distribution result is "Failed"
For the client machine 2 for which the notification of the above was received or the distribution result notification itself could not be received, after removing the cause of the failure, start from step 23 again,
A new version of software is transmitted from the server 1 to each client machine 2.

Embodiment 2. In the first embodiment, all the software is handled uniformly, but depending on the type of software, there are some that can only be supported by the new version, and conversely, the old version can be used without any problem for the time being. Some can be done. It may be more effective to change the handling depending on the importance of software. As the above method, there is a method shown in FIG. 7 as one embodiment of the operation of the system shown in FIG. FIG. 7 shows only the main part. At the time of software distribution from the server machine 1, if at least one client machine fails to be distributed in the network system, the server-side operation setting file determining means 18 analyzes the contents of the server-side operation setting file 17. As shown in FIG. 2 as an example of the server side operation setting file 17, if the server side operation setting file 17 describes a list of software names and y or n indicating whether or not to restore each software, As shown in the flow chart of FIG.
When distribution fails, the operation setting file judgment means on the server side
It is determined by the distributed software whether the software is recovered (step 42), and if it is determined that the software is recovered, the software recovery instruction means 15 is executed.
(Step 43), which causes the client machine 2 to execute the software recovery means 16 (Step 4).
4) Copy all modules of the old version of software to the original space on the disk and restore the software to its original state before distribution (step 45).

Further, in the case of large-capacity software, it may be more efficient to change the handling for each module, and there is the method of FIG. 8 as one embodiment of the operation of the system shown in FIG. FIG. 8 shows only the main part.
As shown in FIG. 2 as an example of the server side operation setting file 17, in the server side operation setting file 17, a list of software names and module names constituting each software, and whether or not to recover those modules y Alternatively, if n is described, as shown in the flowchart of FIG. 8, when the distribution fails, the server side operation setting file judgment means 18 judges whether or not the software is recovered (step 46) If it is determined to restore the software, the module name described as the module to be restored in the server side operation setting file 17 is set in the software restoration instruction (step 47), and the software restoration instruction means 15 is set. Execute (step 48), which causes the client machine 2 to recover the software. Run 16 (step 49), copy all modules of the old version of the software to the original space on the disk, and restore the software to its original state before distribution (step 50). As described above, the behavior when distribution fails can be set for each software and each module, so if there is any client machine that failed distribution in the network system, the effect of the new version However, by restoring the other software and modules to the old version, leaving only the strong ones, it is possible to recover quickly even in the incomplete state. The processing gap can be reduced.

Embodiment 3 In the first and second embodiments described above, the recovery process is performed depending on whether or not even one client machine has failed to be distributed. However, there are tens or hundreds of client machines. If a large number of client machines are connected to the network, it may be more effective for the entire system to decide whether or not to recover depending on the number of client machines that have failed to distribute the software. As the above method, there is a method shown in the flowchart of FIG. 9 as one embodiment of the operation of the system shown in FIG. This method
In step 36 of FIG. 4 in the first embodiment, the sentence “Is there any client machine that failed in distribution?” Is replaced with the sentence “Is the number of client machines that failed distribution outside the allowable range?” This is a modification of the partial processing flow. Before the distribution is started, the allowable range of whether the distribution result judgment standard file 13 of the server machine 1 is judged to be successful when the distribution to a certain number of all client machines 1 is successful is made in advance as a percentage. It is described by the numerical value of. The server machine 1 receives the distribution result notification from the client machine 2 by the distribution result receiving means 11
Is received (step 51) and recorded in the status recording file 12 (step 52). When reception of the distribution result notification from all client machines is completed (step 53) or when a predetermined time has elapsed (step 54), the contents of the distribution result judgment reference file 13 and the status record file are calculated by the distribution result judgment means 14. The contents of 12 are collated (step 55), and if the ratio of the number of client machines that failed distribution is greater than or equal to the value described in distribution result judgment standard file 13,
The software recovery command means 15 is executed (step 5
6). In each client machine 2 that has received the software recovery command (step 57), the software recovery means 16
Thus, the previously stored software of the old version is copied to the original space on the disk, and the software is returned to the original state before distribution (step 58). Steps
If the ratio of the number of client machines 2 that failed in distribution at 51 is less than the numerical value described in the distribution result judgment reference file 13, the process ends as it is. As described above, according to this method, the operation at the time of distribution failure is determined by the ratio of the number of client machines 2 that failed in distribution, and if the number of client machines 2 that failed in distribution is within the allowable range of the system, software recovery means is used. Since 16 is not executed, flexible operation is possible considering the efficiency of the entire system, especially when the number of client machines 2 is large, and the allowable range can be set appropriately according to the operation purpose of the system. it can. Even if the allowable range is exceeded, it is possible to set whether to execute the software recovery means 16 for each software, and set the target software according to the operation purpose of the system. You can

Embodiment 4 In the first to third embodiments described above, the server machine initiative type and the circumstances on the client machine side are ignored because the efficiency of the entire system is emphasized. However, when the number of client machines is small, it is necessary to consider the efficiency of the entire system while considering the circumstances on the client machine side. In order to realize such operation, there is a method shown in the flowchart of FIG. 10 as one embodiment of the operation of the system of FIG. In FIG. 10, by executing the software transmission means, the server machine 1 transmits the packaged version of a plurality of software to all the client machines 2 in the network system (step 59), and each client machine 2 The software is received by the software receiving means 7 (step 60), and the new version software transmitted is stored as a backup in case the expansion fails (step 61). When the software is received from the server machine 1, each client machine 2 analyzes the contents of the client side operation setting file 19 by the client side operation setting file judging means 20 (step 62). In the client-side operation setting file 19 shown in FIG. 3, "software name", "whether or not to immediately update", "whether or not to save the software (module) of the old version at the time of updating", and "update" Information of "module name" is stored. If it is determined in step 56 that the update is to be performed immediately, the contents of the client-side operation setting file 19 are further analyzed (step 63). If the old version software (module) is saved at the time of update, the old version software After saving (step 64), if the saving is not performed, the new version software is immediately developed and replaced with the old version software (step 65). When it is determined in step 62 that the update is not to be performed immediately, it is determined in step 66 whether or not a predetermined time according to the “set time” (not shown) set in the client-side operation setting file 19 has elapsed. If the time has passed, the contents of the client-side operation setting file 19 are further analyzed (step 67), and if the old version software (module) is saved at the time of updating, the old version software is saved (step 68). Later, if not saved, the new version software is immediately expanded and replaced with the old version software (step 69). In step 66,
When the predetermined time according to the “set time” set in the client side operation setting file 19 has not elapsed and the preset time has elapsed from the time of step 59 in the server machine 1 (step 70), The software development command is transmitted from the server machine 1 to all the client machines 2 (step 71), and each client machine 2 receives the software development command (step 72).
Later, if you want to save the software (module) of the old version when updating (step 73), save the software of the old version (step 74), and if you do not save it, immediately deploy the software of the new version. And replace the old version of the software (step 75).

As described above, the time for deploying the new version of the software can be postponed due to the circumstances on the client machine side. Therefore, when the old version software is being used and the old version is to be continued for a while, the process can also be performed regularly. It is possible to deal with the situation where the processing data in the old version has not been processed due to the processing delay.
Further, in the server machine 1, when a preset time has elapsed from the time of transmission, the server machine 1 transmits a software expansion command to all the client machines 2, so while considering the circumstances on the client machine side. It is possible to improve the efficiency of the entire system. Also, since the client-side operation setting file 19 can set the operation for each software, it is possible to expand only the required new version of the software and postpone the time to replace the old version of the software, simplifying the processing content. It is possible to realize a processing oriented to the purpose. Further, whether or not to save the software of the old version can be set for each software. Therefore, when there is little free space in its own disk device 3 due to the circumstances of each client machine 2, the old version with a relatively large capacity can be used. The software can also take steps that do not save. If necessary later, the server machine 1 may request the transmission again.

As described above, the contents of the client side operation setting file 19 can be executed as a processing flow without human intervention, but the operator at the client machine 2 at the time of transmission from the server machine 1 It is also possible to wait for the instructions. By waiting for the operator's instruction, it is possible to deal with the latest situation on the client machine side at the time when the new version of the software is transmitted, so that more effective system operation becomes possible.

[0025]

Since the present invention is constructed as described above, it has the following effects.

When the distribution result judging means judges that the distribution has failed in at least one client machine, the new version of the software distributed from the server machine to all the client machines is regarded as the old version before distribution. Since the software recovery instruction means for instructing the replacement is provided, in a system where it is strictly required that the software of the same version always operates, there is an effect that it is possible to take an emergency measure to restore the old version.

Further, when it is determined that the distribution has failed in at least one client machine, or when a predetermined time that has been set in advance has elapsed since the new version of the software was transmitted, the server machine described above can be sent to all the client machines. On the other hand, the new version of the software that has already been distributed has been replaced with the old version of the software that has not been distributed, so even if the distribution result sent from the client machine to the server machine does not reach the server machine, wait a certain period of time before the old There is an effect that you can take the first aid to return to the version.

Further, depending on the ratio of the number of client machines that failed to be distributed to the total number of client machines, the new version of software that has been distributed to all client machines may be replaced with the old version of software that has not been distributed. Since it decides whether or not to instruct, it is possible to operate flexibly considering the efficiency of the entire system when the number of client machines is large. Can be set appropriately.

Further, since it is possible to set, for each software, whether or not to replace the software of the new version with the software of the old version, it is necessary to leave only the software which cannot be handled unless it is the new version. By restoring the software and modules of the previous version to the previous version, there is an effect that it is possible to recover quickly even in an incomplete state.

Further, since the client machine selects whether to immediately replace the received new version software with the old version software or after a lapse of a predetermined time, the old version software is replaced by the client machine side. There is an effect that you can continue to use.

When a preset time elapses after the new version software is transmitted from the server machine to the plurality of client machines, the new version software is replaced with the old version software from the server machine to all the client machines. Since the command for exchanging is transmitted, there is an effect that the efficiency of the entire system can be achieved while considering the situation on the client machine side.

Further, since the time when the new version of the transmitted software is replaced with the old version of the software can be set for each software depending on the circumstances of the client machine, the processing can be simplified and a flexible response is possible. There is an effect.

Further, since it is possible to select whether or not to save the software of the old version in the client machine, it is possible to take a measure not to save the software of the old version having a relatively large capacity.

[Brief description of the drawings]

FIG. 1 is a block diagram showing a system configuration of a first embodiment of the present invention.

FIG. 2 is a server-side operation setting file according to the first embodiment of the present invention.

FIG. 3 is a client-side operation setting file according to the first embodiment of the present invention.

FIG. 4 is a flowchart showing the operation of the first embodiment of the present invention.

FIG. 5 is a file layout diagram showing a distribution result notification according to the first embodiment of the present invention.

FIG. 6 is a file layout diagram showing a state recording file according to the first embodiment of the present invention.

FIG. 7 is a flowchart showing the operation of the second embodiment of the present invention.

FIG. 8 is a flowchart showing the operation of the second embodiment of the present invention.

FIG. 9 is a flowchart showing the operation of the third embodiment of the present invention.

FIG. 10 is a flowchart showing the operation of the fourth embodiment of the present invention.

FIG. 11 is a system configuration diagram showing a conventional software distribution system.

[Explanation of symbols]

1 server machine, 2 client machine, 4
Software to be distributed 6 Software transmitting means, 7 Software receiving means, 8 Software storing means, 10 Distribution result notifying means, 14 Distribution result judging means, 15 Software recovery command means, 16 Software recovery means.

 ─────────────────────────────────────────────────── ─── Continuation of front page (72) Inventor Yuji Aoki 2-3-3 Marunouchi, Chiyoda-ku, Tokyo Sanryo Electric Co., Ltd.

Claims (8)

[Claims] 1. A network system comprising a plurality of client machines connected to a server machine,
Software saving means for saving the old version of the distribution target software on the distribution list sent from the server machine in the client machine, software sending means for sending the new version of software from the server machine to the client machine, sent Software receiving means for receiving software in the client machine, distribution result notifying means for notifying the server machine of the success or failure of reception, and it is judged whether the distribution to each client machine has succeeded or failed based on the notified distribution result. Means for determining distribution result, and when this means determines that the distribution has failed in at least one client machine, the server machine of the new version already distributed to all the client machines. Software for instructing to replace the software with the old version of software before distribution, and software recovery for returning the old version of the software saved by the software saving means to the original state when the software recovery instruction is received. A software distribution system consisting of means. 2. A step of storing, in a client machine, an old version of distribution target software on a distribution list transmitted from a server machine, a step of transmitting a new version of software from the server machine to the client machine, and the transmitted software. In the client machine, informing the server machine of the success or failure of the reception, determining whether the distribution to each client machine is successful or unsuccessful according to the notified distribution result, the result of this judgment When it is determined that distribution has failed on at least one client machine or when a preset time has passed since the new version of software was sent, all client machines From the step of instructing the software to replace the new version of the software that has been distributed with the old version of the software before distribution, and the step of restoring the old version of the software saved when receiving this command. Software distribution method. 3. In the server machine, when it is determined that the distribution of the new version of software to each client machine has failed, the total number of client machines that have failed to be distributed is determined by the proportion of all the client machines. 3. The software distribution method according to claim 2, wherein it is determined whether or not to instruct the client machine to replace the distributed new version software with the undistributed old version software. 4. The software distribution method according to claim 2, wherein it is possible to set, for each software, whether to replace the transmitted new version of the software with the old version of the software. 5. A step of transmitting a new version of software from a server machine to a plurality of client machines, a step of receiving the transmitted software in the client machine, and a step of immediately replacing the received new version of software with old version of the software. A software distribution method comprising a step of selecting in each client machine whether to replace after a lapse of a predetermined time. 6. When a preset time elapses after transmitting the new version software from the server machine to a plurality of client machines, the new version software is transferred from the server machine to all the client machines to the old version software. 6. The software distribution method according to claim 5, further comprising transmitting an instruction to replace with. 7. The software distribution method according to claim 5, wherein it is possible to set a time for replacing the transmitted new version software with the old version software for each software. 8. The software distribution according to claim 2, wherein it is possible to select whether or not to save the old version software of the new version software sent from the server machine in the client machine. Method.